Editorial notes

NOTES
Bechbretha is not given as the title at the beginning of the text in A, but
nevertheless it was clearly the old title of the tract. In §18 the phrase i mbechbrethaib marks the return to the subject of the tract after the excursus on fidbretha
in §§12-17. In §27 the same phrase characterizes the subject of the tract. In §§36
and 55 (the colophon) it is used for judgements made by brithemain within the
area covered by the tract, judgements which should be in accordance with its
teaching. Breth is always used in the plural in the titles of law-tracts ; cf. Bretha
Comaithchesa , Bretha im Gatta, etc. The title is No. 32 in the list, mostly of
law-tracts, printed by Thurneysen, ZCP xviii 362ff.
§1
This sentence introduces the first topic of Part I of the tract, the tairgille
‘ fore-pledge ’. That the giving of fore-pledges in respect of bees was indeed a
practice which caused difficulty and dispute is shown by other texts. An O. Ir.
text on the law relating to flying animals (Appendix 5 of this edition) takes a
different line from that adopted in BB, denying that a tairgille should be given for
bees ‘ because they are swift . . . and they do not all commit leaping-trespass
together.’ For hens, on the contrary, one should give a fore-pledge as for livestock, since they can be kept in by a fence. The assumption is clear: a fore-pledge
is given by a man to his neighbour when he can and should prevent trespass by
his animals. The fore-pledge guarantees the payment of compensation and fine
for such trespass. But he cannot prevent bees from trespassing, and therefore he
should not have to give a fore-pledge. BB' s point of view, to judge by §§1-10 and
24, is that the fore-pledge guarantees the neighbours that if they allow the new hive
its three years of soire (immunity from claim), they will receive their rightful
dues, in the shape of one swarm each in the next two years. Once they have
received their rightful due, the tairgille has performed its function, and should
be returned. In effect, therefore, BB accepts the argument that a tairgille cannot
be used as for other animals; but it requires the tairgille to guarantee a form of
compensation to the neighbours which is special to bees. The neighbours are
to be assured that, if they do not complain about the bees, they will themselves
acquire their own hives from the swarms of the fourth and fifth years. Then they
will all have hives and the trespasses will cancel out. Though this solution
successfully adapted the main principles of the law of comaithches to the special
case of bees, it was plainly not universally accepted. The text already quoted
takes a different line, and yet another is taken by a text on the forms of distraint
(ed. D. A. Binchy, Celtica x 80 §11 = our Appendix 6), again an O. Ir. text
somewhat later than BB. This text identifies the tairgille with either a payment
of produce of the bees ( torad ) or a swarm given by the owner of the hive to his
neighbour. If such a tairgille is not given, then the owner of the land trespassed
upon by the bees may destroy any he catches. This is a much less satisfactory
line than that taken by BB. The tairgille no longer, as elsewhere, guarantees
compensation; instead it is itself the compensation. This is a violent change to
the meaning of tairgille , whereas BB provides a genuine legal solution. Gubretha
Caratniad §12 ( ZCP xv 319) appears to agree with BB (see note on §10).

§§1,2]
NOTES
91
annsam. For the syntax here, with a superlative opening the sentence and
functioning as a predicate, see Binchy, Erin xviii 49.
For the meaning of annsam see note to §15 of B Crolige (Eriu xii 60). Dr.
Binchy distinguishes three meanings: (a) ‘most difficult to determine’ (e.g.
B Crolige §§30, 38, 52); (b) ‘ most troublesome, most oppressive ’ (e.g. B Crolige
§15); (c) ‘most remarkable, most curious’. Here it has the first meaning.
Similarly, asu in §§36 and 39 means ‘ easier to determine.’
For the spelling in -om of the MS, see Introd. p. 3.
tairgille. Most other cpds. of gell retain the neuter ostem inflexion of the
simplex, e.g . forgall, fuigell, airgell ( aurgell ). Tairgell occurs in CG 241 (see note)
but the normal form of the cpd. to-air-gell- is tairgille [n. io]. Gell is also replaced
by -gille in the cpds. lethgille, langille, ingille, firgille ( Grammar §254).
A tends to spell the word with taur-, §§1, 26 (see Introd. p. 3) but has
tairgille in §24, and targillib in §3. The spelling with tair-, supported by other
MSS in § §1 , 3 and 26 is clearly correct: cf. do-airgella vb. n. tairgel/ad.
§2
alid-side. It is not easy to say what is referred to by side. Alid , ‘ entails, is
entitled to ’, is a verb much used in some law-tracts (e.g. B Crolige 35 exx.), but
not in others (e.g. CG no exx.). Even those tracts which favour the verb use it
differently. In B Crolige, all but seven exx. clearly have a personal subject; and
in some of the others the subject is a word which can be used personally (e.g./w/7
4 wound; wounded man ’). In BB, however, a clear majority have a non-personal
subject. We may compare §§41 and 42. In both cases we have a paragraph which
begins with a nominativus pendens. This serves as a heading which defines the
situation for which the rest of the paragraph will provide a solution. First the
problem is set out — -and sometimes the problem is annsam 4 most difficult ’ — and
then the solution is given (cf. Introduction p. 35). In §41 the problem is beich
thetechtai gaibte barr nemid no maigin no brat scarthae, 4 tracked bees which
settle in the branches of a dignitary or on open land or on a spread cloth ’. The
solution begins dilid nemed 4 the dignitary is entitled . . . ’. Here, then, dilid has
a personal subject. In §42 there is the same pattern of problem and solution. The
problem is set out by a nominativus pendens construction, the solution by a
sentence beginning with dilid-son. Son does not refer back to any particular word
in the statement of the problem : it seems to refer to the problem as such. Similarly
in §46 we have a statement of a problem, fer fo-gaib frith mbech etc. The solution
begins dilid cethramthain a thoraid . . . do fiur fod-gaib. Here again the subject of
dilid seems to be the whole situation stated by the nominativus pendens phrase.
There are some more familiar exx.: in §19 the pronouns ed and on seem to
refer back to tir in §18; in §29 fuil (in a non-personal sense) is the subject ; in §16
the implied subject of ni aili is perhaps fid. In §10 the MS variation between A
and B makes firm conclusions impossible (see note).
There remain §12, alid-side cro ime thorad, and the present ex. In both alid
is followed by the anaphoric pronoun side. They are distinguished from the ex.
in §42 by the use of a masc. rather than the neuter pronoun son. At first sight
the use of the masc. side rather than the neuter son seems to exclude the idea that
the pronoun refers to a situation or a problem and not a person. In §52, also, it
is the neuter son which refers back to a situation previously described. Yet the
difficulty is not insuperable: in §39 the clause ni asu araile introduces a new
problem and araile refers to the whole situation, not to a particular word; and

92
NOTES
[§2
yet the pron. is masc. (cf. CIH 467.6; 474.4, 24). In §12 one might have expected
sdn referring to barr-bunad, or ade referring to breth; side unless it refers to an
unmentioned fid must refer to the situation. Similarly in §2: §1 has stated a
problem; as araile in §39 shows, a masc. pron. may be used to refer to such a
problem and there is no objection to interpreting side in the same way.
Another solution is suggested by Professor Quin. Since the opposition
between final [t], [0], [5] and [d] is neutralized before s (all are realized as [t],
Grammar §§139, 185(d)) both dilid and ailit would have had [t] before side or sidi
(cf., with s followed by d giving -s /-, bes tech §20). One might, therefore, suggest
that alid side is the result of false etymological spelling, and that the original had
alit sidi, sidi referring to the bees mentioned in §1 (cf. dlegait beich §5). While
attractive in isolation this solution takes account of a narrower range of evidence
than the one proposed above and is therefore less satisfactory.
tir bes-da nesam each leth. Cf. CIH 1712.4-6 (normalised): ocus arindi
do-n-airgel/a each ara chethrai a coir comathchesa frisna .iiii. comaithchiu ata
nesam immid-mbiat fri da thoib -] fri da n-airchenn ‘ and it is for this reason that
each man gives a fore-pledge in respect of his livestock, in accordance with the
law of neighbourhood, to the four neighbours who are closest, who surround
him on the two sides and on the two ends [of his own holding].’ The tendency
was to visualize holdings as rectangular blocks lying side by side and end to end.
Neighbours were normally kinsmen; cf. CIH 64.18: Cair: can for-beir comaithches? A ilchomarbus. ‘ A question: whence grows joint-husbandry ? From
plurality of heirs.’ Hence the tendency was to think of the five neighbouring
holdings as the lands of a kindred, all the more so since the terms nessam,
comnessam, comoccus were regularly used of proximity of kinship. Proximity of
holding and proximity of kinship were expected to go together. As a result it
was quite natural for the compiler of BB to treat the divisions of swarms by
neighbours as an instance of the division of kindred-property, fintiu (§§10, 11,
18-22). Examples of cdic t'nreba being used for the lands of a kindred (e.g. Stud.
E. Ir. Law, pp. 148, 149) show that the idea of five adjoining holdings as the
finteda of kinsmen was a standard legal schema. One man was, therefore,
supposed to have four kinsmen-neighbours.
The compiler of BB, however, gets tangled in the ramifications of this idea
just as he does in CU (cf. the introduction to CU, Erin xvii 59-60). His muddle
has been further compounded by subsequent revision. His mistake was to mix
his analogies by not keeping separate two forms of inheritance : first, inheritance
from a man whose land did not pass to a descendant, and, secondly, inheritance
by more remote kin of the lands of a group of close kinsmen who had all died
without issue. In the first case the close kin took all the inherited property of the
dead man; in the second, different kindreds, related more remotely to the extinct
kindred, shared its property between them. The conception of five neighbouring
kinsmen is tied to the idea of a single kindred co-operating in farming; but in
§§18-22, as in CU, the compiler uses the analogy of different but related kindreds
sharing property between them. He changes analogies in mid-stream because he
wishes to give an order of precedence according to which the neighbours received
their swarms: the nearer the neighbour’s land and the better the grazing it
provided, the sooner was a swarm received. He cannot extract such an order of
precedence from the first analogy because in such a case the kinsmen took equal
shares at the same time.
CU and BB both make attempts to lessen the confusion. CU has only three
neighbouring holdings so that they can then be identified with the derbfine,

§§2, 3]
NOTES
93
iarfine and indfine (used for close kin, remoter kin and most distant kin). BB
has four neighbouring holdings, but abandons the second analogy after
mentioning the derbfine in §18. The insertion of gelfine in §11 appears to be a
later addition (see note to §18).
ciapa meit ciapa laget. On the difficulty of analysing forms of the indefinite
pronoun, ce, ci, da, followed by the copula or substantive verb, see Grammar,
§458; Bergin, Erin xii 208 and DIL s.v. 1 da, B. Ciapa is probably da (weakly
stressed: class (a) of Grammar, §458) -f -p, 3 sg. pres. subj. of the copula, + a
3 sg. possessive pronoun; cf., with subst. verb rather than copula, Ml. 61b28,
cia-be a mmet. The analysis is supported by the following exx. from BFG: dapa
lin dpa n-uaite ( CIH 470.6); ciaba lin cipa n-u[a]ite (470.23); da lin cipa n-uaite
(470.31); daba lin ciba n-uaite (471.11). The nasalisation suggests that the analysis
is d(a)p an ‘ whatever be their Cf. dapa lin dapa n-uaite, B Crolige §57.
Elsewhere the indefinite pronoun appears before the 3 sg. pres. subj. of the copula
in the reduced form ci; for exx. see DIL s.v. 1 da, B. We have, however, hesitated
to emend to cipa. While BFG always has cipe before a relative verb ( CIH 467.9 ;
476.28) and dp before a noun ( CIH 474.13), it usually has the fuller form da
in the present construction (9 exx.) as opposed to ci (4 exx.): CIH 465.21-22;
466.18; 467.38; 470.6, 23, 31; 471.11. Cf. ciapa lin ciapa n-uaite, B Crolige §57,
but dp magen §40, dp da §44.
For exx. of the idiomatic use of meit and laget see DIL, s.v. lagat. The quality
of the g in laget fluctuates: the only ex. from the glosses, Sg. 26all, has a broad
g; but later MSS tend to show a palatal consonant. For the related comparative
of becc, laugu/ luguj laigiu, the glosses show forms with both broad and slender
g; for exx. see DIL, s.v. bee.
§3
a tairgillib. We translate ‘ according to fore-pledges ’ since we understand
the tairgille to be a pledge given at the outset guaranteeing the payment of a
swarm to each neighbour after the end of the three years of immunity. Dr. Binchy
points out that it could be translated ‘ out of fore-pledges ’ and taken to imply
that payments for tairsce etc. were subtracted from the tairgille. This interpretation finds some support from the text on distraint, Appendix 6 (see general
note on §1) and from the B gloss a to this paragraph which would have the value
of the tairgille vary according to the ‘ nearness and contiguity of the grazing ’.
The value of the tairgille would then vary according to the expected penalties
for tairsce etc. It would not be a good argument against this interpretation to
point to the third glossator’s ./. geall da screpall, ‘ i.e. a pledge of two scruples ’
in gloss b to §1 (similarly the glossator in C, and the first glossator in gloss a to
§2). Two scruples would be an inadequate sum from which to subtract several
fines; however, the much earlier B gloss a assumes, as we have seen, that the
value of the tairgille can vary. The later glossators appear to have thought that
the fore-pledge for bees was the same as the ordinary fore-pledge given according
to the rules of comaithches which they also believed to be two scruples (cf. CIH
64.33-4, gloss on each tairgilli; similarly 64.24-6). In this belief, however, they
must be wrong since the fore-pledge for bees only operated until the neighbours
had received their swarms in the fourth and fifth years, while the fore-pledge of
normal comaithches guaranteed a relationship of indefinite duration. Whereas
BB §24 envisages the return of the tairgille, B Comaithchesa appears to assume
that neighbours will keep the fore-pledges exchanged at the initiation of their

94
NOTES
[§3
relationship. The evidence of the glossators, therefore, on the value of the
tairgille in BB must be rejected. On the other hand, B Comaithchesa itself lends
no support to the theory that penalties might be paid out of the tairgille. A
permanent tairgille would probably consist of some durable commodity such as
the articles of silver or bronze or yew mentioned in CG 282-3, whereas the
penalties of B Comaithchesa consist of sacks of grain ( CIH 66.321T.), or a proportion of rent (69.14-5; 71.23; the rent itself consisting of a bovine animal:
70.19, 26), or the temporary alienation of pasture (70.34; 71.14-5; but this may
be the colainn feich as opposed to the smacht: 75.2). In BB , the smachta seem to
be identical with the gratis payments of swarms to the neighbours (cf. § §7, 9
and 19-22). It is probable, therefore, that the tairgille of B Comaithchesa and the
tairgille of BB were given to guarantee compensation and not to provide compensatory penalties should they be due.
tairsce. Tairsce is a type of trespass defined in B Comaithchesa ( CIH 75.15)
as ‘ coming across a holding or across two ; also going across a road, going across
a river which it is not necessary to swim, crossing over the deserted land of an
absentee ’. It is distinguished, as the lesser of two trespasses, from ruiriad
‘ stampede ’, ro + reth {CIH 77.38). Both tairsce and ruiriud seem to be used for
trespass over unfenced land, as opposed to airlim, ‘ leaping ’, and feis , ‘ spending
the night which are the corresponding two grades of trespass in enclosed land.
Here again BB and the later text on caithchi bech (Appendix 5) adopt different
views. For BB trespass by bees is tairsce , presumably because fences are no
obstacle to bees. For the other text, it is airlim since, where there are no fences,
animals should be prevented from trespassing {tairsce or ruiriud ) by a herdsman;
but bees are too swift and fly off at different times, so that no man could possibly
prevent them from trespassing. If neglect by a herdsman is not in question,
it cannot be tairsce, and if it is not tairsce it must be airlim. The difficulty is
caused by the refusal of the lawyers to abandon the traditional terminology of
the law of comaithches, even when the implied distinctions are wholly out of place.
gloss 3a B. dr tarrailf. We are uncertain of the meaning of this gloss on
tairsce. There are two possible explanations. First, tarraill may be for taraill,
3 sg. perf. of do-aidlea 4 visits ’, i.e. 4 land which it (the bee) has visited ’. This
gloss might then have to be taken as later than the rest of B’s glossa, as it has a
singular subject (‘ the bee ’ ?) whereas the preceding verb has a plural subject
(‘ the bees ’)• Alternatively, one could read tir tar aill 4 land across another
[land]’. B Comaithchesa understood tairsce as including the preposition tar,
giving the definition tuidecht tar seilb no tar a di 4 coming across a holding or
across two ’ {CIH 75.15; cf. the corresponding etymological definition of ruiriud
as rith tar teoru selbu no cetheoru selbu, CIH 77.38, and the later glosses, CIH
198.13, 15). In BB, the working assumption is that the bee-keeper is surrounded
by four immediately adjacent holdings ( § §2, 3). Hence the very definition of
tairsce may have raised a problem in the glossator’s mind in that it suggests the
possibility that not only the bee-keeper’s immediate neighbours but also their
neighbours might be entitled to compensation. In a discussion of tairsce,
probably of the late O. Ir. period, the maximum group of neighbours among
whom there can be imuaim folad 4 mutual binding of obligations ’ consists of
the four immediate neighbours and the four neighbours nearest to them {CIH
198.26-8; the corresponding text CIH 579.23-4 contains an omission). Tir tar
aill might therefore refer to the neighbour’s neighbour’s holding mentioned in
B Comaithchesa' s definition of tairsce.

§§3,4]
NOTES
95
i cinaid. Cin appears to refer, in BB , to personal injuries committed by bees;
cf. §§29, 31, 33, 34. In the text on the forms of distraint (Appendix 6) it seems to
be used of grazing by bees on another man’s land: cin a mbel is identified with
orgain do bechaib, ‘ damage by bees ’, which is one of the three injuries to land
most difficult to assess ( ar at he teora fogla ata annsam file do thir orgain do
bechaib -] echaib *j mucaib). Cin is the most general word in Irish law for an offence
and thus its particular application is at the will of the compiler. A’s gloss imcin
ccechad is on the right lines but too restricted, an example not a definition.
i lloge. In BB log is only found here and in §4. Its meaning is value, worth
or price considered from the point of view of exchange, whether by sale, gift or
compensation. If fair see and cin both refer to offences by the bees (against the
neighbour’s land and against persons respectively), then log seems to be out of
place. Even if it refers here to compensation paid for some offence committed
by bees, this will not distinguish it from the payments in respect of tairsce and
cin. In §4 loge appears to refer to the payments of swarms to the neighbours
which are due in the fourth and fifth years after the first acquisition of a hive, the
deolaid referred to in § §4, 9 and 19. It is apparent from these later sections that
the neighbours have a claim against the owner of the hive because his bees
trespass on their land while grazing. This claim cannot be pursued during the
three years of immunity, soire, enjoyed by a new hive, but becomes operative in
the fourth and fifth years. The basis of the claim is, therefore, twofold : (a) the
bees commit tairsce, (b) the bees graze the neighbours’ land while committing
tairsce. It is this fine distinction upon which the glossator in B seizes : the value
of the tairgille should correspond to the extent to which the bees graze a particular
neighbour’s land. If, then, the land is close to the hive and thus heavily grazed,
the tairgille should be that much greater (B’s gloss a). B’s gloss d then draws the
further conclusion that, as a large tairgille implies heavy grazing, so it should
imply a smaller price to be paid by the neighbour to the owner of the hive for a
swarm. B’s argument here contradicts the text, for the latter lays down the rule
that the swarm given to the neighbour by the owner of the hive is a deolaid, a
gratis gift for which no payment, large or small, is due. Though B’s explanation
cannot be entirely correct, it may have caught the meaning of the text insofar as
it distinguishes between the claim arising out of trespass and the claim arising out
of the value of the grazing, for the text itself appeals to the latter criterion in
§§18-22. This interpretation accords with the use of the plural, loge, both here
and in §4, for the four lands mentioned in §3 may, according to §§18-22, afford
grazing of different values. Log, then, seems to refer to the swarms given gratis
insofar as they compensate for the bees’ grazing rather than their trespass.
ar chethardoit tire. The conception of four neighbouring holdings has already
appeared in §2. The phrase cethardoit tire occurs here and in §9. Cethardoit is
otherwise unattested; but the word appears to be a collective, meaning ‘ group of
four ’. Cethar- is plainly the compositional form of the numeral cethair, and -doit
appears to have been abstracted from trindoit ‘ trinity ’. In an etymological gloss
the A glossator equates -doit with doicl (earlier doe ) ‘ arm ’, but this explanation
does not provide a suitable meaning for cethardoit.
§4
dlegait-sidi. Both sidi and side are found in the Glosses for the nom. pi. of
side, although sidi is more common, and must be the older form.

96
NOTES
[§§4, 5
deolaith. This word is written deolicl , Wb. 2cl9, and the plural is written
deuladte in the glosses on St. Augustine’s Enchiridion ( ZCP vii 485), in both
cases without an accent on the first syllable. It also occurs as deolaith in Gubretha
Caratniad §12 (ed. Thurneysen, ZCP xv 319). Thurneysen, Grammar, §831,
analyses the word as di f folad and marks the e long. Greene, Erin xxvii 37,
suggests that syncope in such forms took place before the loss of intervocalic w
and thus the sequence *-ewa- was transformed into the long diphthong -eubefore -w- was lost. For deolaith this would presuppose a development
*dewolotis > *dewaldtis > *deuloth > deolaithldeulaid (cf. deuladte). This is
certainly possible, but it assumes that deolaith is an old compound, which may
well not be the case. (In any event the preposition di- must have the by-form desince * di-wo - regularly yields du- \ Greene, p. 34). It is also possible that deolaith
is a post-syncope formation or that it has been reformed by analogy with the
simplex folad. If either of these suppositions were correct, one would expect a
hiatus form with a short vowel in O. Ir. deolaith, later deolaid. This would account
for the attested forms.
§5
The bee-keeper does not have to give any of his bees’ produce to his
neighbours for the first three years, with certain exceptions (§6). This three-year
period is common in BB (e.g. §§13, 17, 37, 44, 45). The author seeks to relate the
three-year period of immunity to the bees’ life-cycle, but there seems to be no
scientific basis for this approach. See below.
The text is not explicit as to whether the three years of immunity are enjoyed
by a new hive or a new bee-keeper. If a man already has bees, does he, for
example, start a new period of three years for each swarm that he hives and retains
in his possession? Practical considerations suggest that it is the first hive of the
new bee-keeper that enjoys the period of immunity. The immunity is from claims
arising out of trespass {tairsce §3, or cun § §7, 8, 38). The trespass is the grazing
by bees of the lands of neighbours (§8). Yet it would be very difficult to distinguish the bees of different hives belonging to the one owner. As Appendix 6
shows, one early Irish lawyer was prepared to contemplate such a procedure,
but his ideas are unclear in detail, possibly unrealistic, and find no echo in BB
itself. It might reasonably be objected that BB just assumes that adjacent holdings
will be grazed and holdings not adjacent will not (§§3, 4 etc.). Even if this be
granted, one is still left with the further argument that once a neighbour has
received his ‘ due ’ or ‘ fine ’ in the shape of a swarm (§§18ff.) he is not entitled
to receive any further smachta ‘ fines ’ (§24). As the B glossator reasonably
implies, this is because the neighbour’s bees will now be grazing the land of the
original bee-keeper so that trespasses should cancel out (§24, gl.d). One would
expect this to apply to all the hives of the neighbours, new and old, for this would
be by far the simplest solution. It might also be objected that BB speaks of bees
being entitled to three years of immunity (§§4, 5), not of a bee-keeper being thus
entitled, but this proves nothing: the bees in question may be those of a new beekeeper and, furthermore, BB slips easily into talking of land when it means the
owner of the land (§§12, 13, 24, 37). Finally, the whole point of the procedure
appears to be that it allows a new bee-keeper to get well established in the three
years of immunity, but gives his neighbours the right to receive a swarm each
subsequently and then prevents them from making any further claim. To allow
any new hive to start the process all over again would be to produce chaos out
of a neat and satisfying scheme.

§5]
NOTES
97
lasa n-aithgenatar. As gu aith no gu eadha (in glossb) is the regular gloss on
a(i)th-, we emend aidgenetar to aithgenatar, cf. §6 aithgeniter A (read -atar).
We take -aithgenatar to be 3 pi. pres. subj. prototonic of ath-gainethar
{faith-gain-, VKG ii §735). The basic meaning of ath-gainethar (see DIL s.v.
ad-gainethar ) is ‘ is re-born.’ In Ml. 66b6 ad-gainemmar glosses regenerations,
and in the Laws it is used of the growing-again of grass, and of the restoration of
loans and levies. In our context ath-gainethar clearly refers to the establishment
of a colony of bees by a person who is starting off bee-keeping. He presumably
acquired a swarm from some other bee-keeper, or found a stray swarm on his
land. The author of BB seems to regard the emergence of a swarm as the 4 birth ’
of a colony and its hiving as its 4 re-birth ’ — hence his use here of the preverb
a{i)th~. One can compare Welsh modrydaf lit. 4 mother of a colony of bees ’
which is sometimes used of an established colony capable of producing swarms
(see Appendix 7 p. 202).
bliadain a tuisten. 4 The year of their origin ’ is presumably the year in which
a swarm is introduced into a new hive. Swarming usually takes place in the
months of May or June.
bliadain a seoil. The word seol (apparently a neuter o-stem) is not given in
DIL and is to be distinguished from seol 4 sail, covering, bed, etc.’ It occurs also
in AL v 352.18 = CIH 1377.38; 1921.24 where the false judgement of a brithem
causes seol netha -j blechta i mesa, and in AL iv 52.8 — CIH 219.18; 1866.10
where the injustice of a king causes disce mb/echta, mi/lead measa, seol neatha.
From these contexts it is clear that seol means 4 scarcity, dearth ’ which is
confirmed by glosses, e.g. CIH 809.5 seol J. caile no gan[n\ugud, ut est seol
neatha\ 1866.11 seol netha J. beg don arbnr. The author of BB seems to be
mistaken in believing that the second year of a hive is one in which there is
normally a scarcity of bees. If the weather is reasonably good, a bee-keeper
could expect his hive to be well stocked with bees in its second year.
bliadain a sil. The third year is called 4 the year of their multiplying ’, i.e. the
year in which the hive sends out swarms. Sil 4 seed, progeny ’ is used elsewhere
of swarms. Thus in a text on the forms of distraint (Appendix 6) one of the two
possible fore-pledges given by a bee-keeper to his neighbour is saithe do do sil
bech 4 a swarm to him of the progeny of bees.’ In a tract in H.3.18 p. 394b ( CIH
920.34), which lists the somuine 4 interest, return ’ of various domestic animals,
the profits of bees include saithe do sil bech 4 a swarm of the progeny of bees.’
The author’s description of the third year as the 4 year of their multiplying ’
does not agree with the general experience of bee-keepers — medieval or modern.
A colony will normally send out a swarm or swarms in its second year, and may
even do so in its first year. The strain of honeybee kept by the Early Irish was
doubtless the 4 British Brown bee ’ {Apis mellifera mellifera var. lehzeni), which,
owing to the depredations of the 4 Isle of Wight disease ’ in 1909-17 was largely
replaced in Britain and Ireland by imported Italian and other varieties. According
to Butler ( The World of the Honeybee p. 21) the British Brown bee was reputed
to swarm frequently. This characteristic is confirmed by the Welsh law-texts
from about the 13th century (Appendix 7). These texts assess the value of 4 the
first [secondary] swarm ’ that comes from 4 the first [primary] swarm ’ and 4 the
first swarm that comes from the bull (second) swarm ’. This shows that 13th
century Welsh honeybee colonies sometimes sent out swarms in their first season.

98
NOTES
[§§5, 6
The evidence from Early Modern Irish legal commentary (c. 12th- 13th
century) is more obscure. Commentary in H.3.18 (Appendix 4 (c)) refers to the
‘ son ’ {mac) and ‘ grandson ’ (ua) of the first swarm, and the ‘ son ’ of the bull
(second) swarm. This passage is preserved in two versions and is best illustrated
diagrammatically :
in cheis fein ‘ the hive itself ’
cetsaithe ‘ first swarm ’ tarbsaithe ‘ bull swarm ’
(= tanaise ‘ second swarm ’)
iarsaithe ‘ after-swarm ’
(= smeraighe
‘ foolish swarm ’)
mac son
mac son
(i.e. secondary swarm)
ua ‘ grandson ’
(i.e. tertiary swarm)
It seems that the commentators envisage that three generations (‘ first swarm
‘ son ‘ grandson ’) can emerge in the same season. However, we are informed
by bee-keepers that this would be most unlikely in Irish conditions. A secondary
swarm might emerge from the first swarm in a good summer, but this secondary
swarm would hardly itself produce a further swarm in the same season.
§6
These three cases where the three-year period of immunity does not apply
are also mentioned in §25.
a ccain chuisc. The phrase cain chuisc ‘ due of punishment ’ does not occur
outside BB. Presumably it means that if a person is injured by his neighbour’s
bees, he is entitled to receive compensation in honey, even during the three years
of immunity. One can compare §29, where a man who is stung by a bee is
entitled to ‘ his sufficiency of honey ’.
The first glossator explains cain cuisc as riagail na dighe cuisg ‘ the rule of the
[thirst-] quenching drink.’ The term deoch cuisc (see DIL s.v. cosc (c)) is used of
a thirst-quenching drink, particularly for an invalid {CIH 421.16 gloss) or old
person (BNnE 315 §11). We have come across no other cases of it being used of
honey (or mead), but commentary on Cain Lanamna {CIH 177.18) says that a
divorced wife is entitled to domna coisc di du mil no saithe beach matrollae
‘ enough for a thirst-quenching [drink] of honey for her or a swarm of bees if
it has emerged.’
According to the first A glossator the deoch cuisg consists of what is left over
(perhaps after the extracted honey has been put in containers) and is not a fixed

§6]
NOTES
99
proportion. The third glossator, however, sets out a scheme in which the deoch
cuisg is related to the quantity of honey obtained. Thus, if the largest size of
vessel is filled with honey, the deoch cuisg is a half-cup, etc. This scheme is
doubtless a glossator’s fiction, and bears no relation to actual practise at any
period.
gloss b III. lestar lulaice, etc. We take lestar lulaice to mean ‘ a vessel in which
a milch-cow (l(a)ulgach) can be cooked.’ This term (also lestar samaisce, lestar
colpthaige, etc.) is found in other commentaries (e.g. CIH 920.32; 2104.32) but
not in the text of the laws. However, in the text of Cain Aicillne ( CIH 483.30)
we find caire lulaice lan do lemnacht 4 a milch-cow cauldron full of new milk.’
Thurneysen ( ZCP xiv 358) takes caire lulaice to mean here ‘ a cauldron in which
a milch-cow can be cooked ’; cf. CG 197 tdithi caire umai hi tallaii) tore 4 he has
a copper cauldron in which there is room for a boar ’, 405 caire i talla boin co
tinniu 4 a cauldron in which there is room for a cow and a flitch of bacon ’.
no. We emend na to no, cf. §25 coin chuiscc no mian ngalair no allabruig
n-ai.
allabrig n-ai. We have no explanation for the second exception allabrig n-ai,
a phrase which does not occur outside BB and its glosses and commentary. The
original spelling of the phrase is uncertain. Here, the main scribe wrote allabrig
naie, which was subsequently changed (either by himself or by a later scribe) to
allabraig naice. In B the two words are quoted separately in the spellings allabrig
and aie. The glosses in both A and B indicate original -brig and ai. In §25 A
(the only MS) has allabruig nai. Spellings in A Commentary (App. 1 (a), (b)) are
ollabruig nai, ollobraigh nai, allabrug nee, allobrugh nee and allabrigh nee. C
Commentary (App. 2 (a)) has ollbrughe noe and oUbrugh nee.
Allabrig. The B glossator (probably working in the 9th century) explains
allabrig as .i. ala brig 4 i.e. another meaning ’, taking alia- to mean ‘ second,
other ’. However, in compounds this ordinal is only attested in the forms all-,
ala-, aili-, never alia- ( Grammar §§394, 488). We have no suggestions for -brig,
taken by the glossators of both A and B to be the fern, a-stem brig 4 force, power,
value, meaning.’ The nasalisation of ai, however, shows that the preceding word
is neuter.
n-ai. We emend aiee A, aie B to ai, which is the reading of A in §25, and is
indicated by the glosses of both MSS ( uman ai A, is si ind ai B). Moreover, the
spelling aie occurs as a variant of ai ( ae ) in other texts, e.g. ZCP xii 364.9.
On the analogy of cain chuisc and mian ngalair one would expect ai to be a
gen. sg. The B explanation is si ind ai a timgaire 4 the ai is the demand ’ indicates
that it was taken to be de (ai) 4 law-suit, litigation ’ which has an identical gen.
sg. This seems more likely than the suggestion by the third A glossator that the
word is ai 4 poetical inspiration, poetry, etc.’ as it has gen. sg. uad or uath in early
material (e.g. Erin xxvi 83 §12).
mian ngalair. There are many cases in Irish law where rules can be waived
in times of stress, danger, sickness, etc. Thus an invalid (presumably from the
four neighbouring lands) is entitled to an unspecified amount of honey, even
during the three years of immunity. The phrase mian ngalair ‘ the desire of [one
in] sickness ’ occurs also in B Crolige §24 (Erin xii 20), cf. mir mein(n) 4 a desired
morsel ’ (for an invalid) CIH 242.17; 387.30.
The use of honey in Early Irish medicine seems to have been limited to certain
types of illness. Thus in §25 of B Crolige (repeated in Triads of Ireland 184) honey

100
NOTES
[§§6-8
is included among the three condiments which are forbidden in the rule of
nursing icain othrusa). The explanation given is: conalai mil broinn i mbi lir
‘ honey disturbs the stomach in which there is looseness of the bowels.’ (In the
Carlsruhe Beda 35a2, Thes. ii 24, Hr glosses colera rubea , a stomach infection
causing severe vomiting and diarrhoea. As honey is a substance of quite high
acidity — see Honey , a comprehensive survey ed. Dr. Eva Crane pp. 168-70 — it
would aggravate a stomach infection). In §45 of B Crolige, on the other hand,
honey is included among the three condiments to which an aire ard is entitled,
except when it is prohibited by a leech.
The medical properties of honey have been highly regarded throughout the
world for thousands of years {op. cit. pp. 453-88). The earliest known written
reference to honey is in a medical prescription on a Sumerian clay tablet of c.
2000 B.C. Honey has antiseptic effects, and is of use in preventing infection of
wounds, burns, etc. Its consumption is also of help in restoring the strength of
an invalid as it contains energy-giving carbohydrates. For a full treatment of
its antiseptic and nutritive properties, see op. cit. pp. 258-66, 388-9.
§7
otha. There is uncertainty about the case taken by the preposition otha
{6 + ta) in O. Ir. According to DIL it is followed by the nom. in O. Ir. and by
the acc. or dat. in Mid. Ir. In law-texts it seems normally to take the dative
e.g. otha cetgradaib CIH 466.22, otha .x. mbl-naib 2300.27; otha suidiu 469.30;
470.6; 471.10; 473.21 etc. On the other hand, there is MS evidence for the use
of the acc. e.g. otha airig nard CIH 2286.33 with nasalisation after airig, cf.
Grammar p. 684 (translators’ note 169).
Other cases are ambiguous e.g. otha leth nun[ge ] CIH 475.26 where the
nasalisation of unge merely indicates that leth is either nom. or acc. A passage
in the Additamenta to the Life of Patrick {Thes. ii 238 = Patric. Texts 172 §8)
also contains ambiguous examples, all with the fern, z-stem glais. These are Thes.
ii 238.6 otha glais telchae (nom. acc. or dat.); 238.7 otha glais conacolto (nom. or
acc.); 238.8 otha crich drommo (nom. or dat.).
The examples of otha in the O. Ir. Glosses do not provide evidence one way
or the other. Consequently in our reconstructed text of BB we follow the MS
readings: §§7, 9, 38 otha inna teora bliadnai {ot{h)a na teora bliadna MS), 29 otha
suidiu {ota suidiu MS).
ind n-oircne. The ‘ limit of trespass ’ of the bees is defined in §8. We read
for ind n-oircne with B, against for rind oirggne A. This reading is confirmed by
the gloss foro roichid ind a noirgne A.
fo-ceirt. The MS has focert which we emend to forceirt as at Wb. 32cll (for
the spelling of -rd as -rt in O. Ir. see Grammar §31). Dr. Binchy tells us that
Thurneysen suggested reading focerdatar here, as a passive form seems to be
indicated by the glossator’s chuirtir.
smachtu. For smacht see notes to § §24 and 26.
§8
ind n-oircne. Thurneysen (ZCP xvi 267) translates ind here as ‘ das Geringste ’
(the least amount). This may well be its meaning in the old rule CIH 336.30-31
ind i n-araill erenar, which Dr. Binchy would translate ‘ a minimum [penalty]

§§8, 10]
NOTES
IOI
is paid for another (minimum offence)’. However, the normal meaning of ind
[n.o] k end, limit ’ fits the BB context better. The text seems to be saying that for
legal purposes a bee is regarded as collecting food as far from the hive as a (loose)
cow would graze in one day. One can compare Audacht Morainn §25 Is tre fir
flathemon . rosaig cech bo cenn a hingelte ‘ it is through the justice of the
ruler .... that each cow reaches the end of its grazing-land ’ (i.e. gets enough to
eat). Here cenn corresponds to ind of BB.
For a similar rough measurement of area, cf. §46 where the extent of the
faithche is said to be ‘ as far as the sound of a bell or the crowing of a cock
reaches.’ According to Dr. Butler, bees generally collect their food within a
radius of about a mile from the hive, (though they sometimes forage as far as
three miles away). It seems likely that a cow would also reach about a mile if
allowed to graze unchecked for one day.
For orcun used of ‘ grazing-trespass ’ by bees, cf. BB §38 and Appendix 6 (c)
or gain do bechaib.
ro-saig. This word is absent in the text, but has been added in the margin
by the third glossator. It must have been in the original as it is glossed (./.
roindsaigidh ) by the first glossator (the bulk of his glosses are older than the
MS — see Introd. p. 15). It is also required by the sense of the passage. For another
textual insertion by the third glossator, see §43 do fiur and note.
tecmallad. We emend tecclamad to tecmallad, as this is the form attested in
O. Ir. (Wb. 14al, 16c25, 16c26). It is vb.n. of do-ecmalla [*to-in-com-ell-]
‘ collects.’
§10
alith. The reading in A, ailit, is also possible. In BB, however, alid is not used
in the plural to mean ‘ they are entitled ’. It is used in the singular and usually
with a non-personal subject (see note on §2). Furthermore, the gloss in B supports
alith. In the text alith has no direct object, though elsewhere it is always transitive.
The gloss in B must be correct in taking deolaithe or deolaid to be the implied
object, and it may be that the text has dropped deolith by homoeoteleuton with
alith. But it is clear from the first part of B’s glossa that in his text alith had no
direct object since his immediate gloss is .i. deolaithe doib etc. It is only in the
second part of the gloss, on a cobdailib crich, that he has ailith deolaid. Hence
the dropping of deolith would have to be early. An early date is also suggested
by the form deolith required for homoeoteleuton with alith , and by its absence
from both MSS. In spite of these difficulties the emendation to alith deolith
remains attractive.
a cobdailib crich. The gloss in B is clearly correct in explaining this phrase
by reference to §§18-22.
ar ni fulaing nech deolaid dialailiu la Feniu. This maxim is also quoted in
Gubretha Caratniad §12 (ZCP xv 319) where Caratnia uses it to support his
judgement that a tairgille should be given in respect of bees. Here it is used to
support the rule that the neighbours are each entitled to a swarm after the three
years of immunity. The two texts are not in conflict since one of the things
guaranteed by the tairgille was the giving of a swarm gratis. The grazing of bees

102
NOTES
[§§io, 11
on neighbours’ land is, therefore, not deolaid , but requires a counter-gift or
payment. But this payment is a gratis gift since it requires no further payment
(cf. note on §4).
ar is si trisfintiu inso. The three are given in §11. For the point of fintiu see
note on tir bes-da nesam each leth, §2.
§11
cobdaili saithe ~j bretha barr-bunaid i usci thairidne tar cricha. Cobdaili
saithe is the main topic of this section of BB, bretha barr-bunaid refers to the
excursus §§12-17 but perhaps also to a lost tract on fidbretha\ cf. note on §14 a
besaib fidbreth. The ownership of trees is also important later in BB since swarms
usually settle in or on a tree. Neither in the excursus, §§12-17, nor later in BB
is there any appeal to the metaphor of sharing by kindreds when discussing trees.
It therefore seems likely that the metaphor was used in the lost tract, and that the
reference here is in the first place to that tract, from which the material in the
excursus may derive. Usci thairidne tar cricha refers to CU. Usci is gen. sg.
depending on bretha.
Barr-bunad here and in §12 we take to be a dvandva compound: see Binchy
in Indo-Celtica, ed. Pilch and Thurow, pp. 38-41 (where he compares the title
in a late list of the qualifications of a judge, i mbechbrethaib itir bun i ban , ZCP
xviii 363 §32). The reading of A here is bairr bunaid. In §12 A has bairr bunaid
at the beginning of the paragraph but barr bunaid at the end. In the latter case
the compound is the subject of a singular verb, dilsigithir {-igthir A).
The situation envisaged in §§12 ff. is of a tree standing by a boundary. Its
base is in one holding but some of its branches overhang the adjoining holding.
The issue is the ownership of fruit falling onto the adjoining holding. Barr
denotes the branches. Bunad, i.e. the place where the bun, the trunk, ‘ sits ’
( <j*bunosedon , cf. Med. W. bonhed ), is the ground from which the trunk grows.
Later in BB, however, bunad has a wider meaning, approximately ‘ origin ’. A
swarm has a bunad : the hive from which it has flown. The owner of the hive is
then fer in bunaid (§38) or fer bunaid (§37) or fer bes bunadach doib (§§37, 41).
The hive itself is bunadach doib (§43). If the swarm passes into the possession
of the man on whose land it settles, his land will then become its bunad, i.e. the
bunad of any swarms it may subsequently send out (§44). The cognate Welsh
term could be used in a similar way as in WML 81.1 : Bonhed gwenyn o paradwys
pan yw, ‘ the origin of bees is from paradise \
gelfine *] derbfine, iarfine *i indfine. For the most recent discussion of these
terms see Binchy, ‘ Irish History and Irish Law: II ’, Stadia Hibernica xvi (1976),
31-38. For their use in BB see notes to §18.
hi fintedaib griain. This spelling of the gen. sg. of the n. o-stem grian also
occurs in §18 and elsewhere in the laws (e.g. Irisches Recht 65 = Fuidir tract §7).
Thurneysen’s rule, Grammar §105, is as follows: ‘ unstressed a between i and a
palatal consonant is narrowed only to e .’ Forms with / or a instead of e (e.g.
liic or liaic for lieic a.sg., d.sg. of lie, lia ‘ stone ’) he regards as later. Thus, for
Thurneysen, the rule for unstressed vowels between consonants (altogether
dependent on the quality of the flanking consonants. Grammar §101) cannot be
universally applied to heterosyllabic sequences of the form VVC. Otherwise one
would expect O. Ir. : *griin . We have hesitated, however, to restore griein, even

§§H, 12]
NOTES
103
though the latter form seems to be attested in O. Ir. (Thes. ii 241.1 = Patric.
Texts 176.3). It is not certain that there is in this case one single sequence of
universal sound-changes: so much is suggested already by the co-existence of
forms such as liic and liaic in MSS subsequent to the O. Ir. period. Later forms
such as grin may descend from *griin which itself developed from *grisani
according to the sequence normal for unstressed vowels between consonants.
Another difficulty is that it is unclear whether raising of unstressed a occurred
before old final -i or -1. Stressed a was not raised (except later before palatal
consonants and those only in special contexts, Grammar §80). But evidence for
unstressed a has been effectively obliterated by the general weakening of unstressed vowels. Thus it is not clear whether the e in griein represents an unstressed a raised to e and preserved only in hiatus (elsewhere >>[a]). A further
difficulty is the possibility of influence from the type of dissimilation found in
aig/ego , saigid/segait ( Grammar §83). Since the syncopation of ia yields e there
are such pairs as liaig/lego (both CG ): if lieig is indeed the older form {Thes. ii
24.38) then liaig may be an instance of this tendency to dissimilation (cf. diaid ,
Thes. ii 340.3, but deid Wb. 13d9, Sg. 212al3, in the u-stem inflexion of dead:
W. diwedd; iairn, Thes. ii 249.6, but iern, Thes. i 2.15; iain n.du. of i'an f. ff-stem,
CG 184). For these reasons the emendation of griain to griein seems to us to be
unjustifiable.
§12
The general problem in §§12-13 is that of sharing the fruit of a tree some of
whose branches overhang a neighbouring holding. On the face of it the solution
is highly favourable to the neighbour. The distribution of the fruit runs on a
cycle of four years: for three years they share it equally between them (§13), and
in the fourth year all the fruit goes to the neighbouring holding (§12). This
suggests that the fruit in question is only that part of the tree’s crop which falls
on the neighbour’s land. There can be no dispute as to the rest. The solution,
then, gives a certain preference to the claim arising out of ownership of the land
onto which the fruit has fallen, but still gives some weight to the claim arising
out of ownership of the tree. It is clear that barr-bunad is used in §12 to denote
the tree under its two relevant aspects : the branches (which partly overhang the
neighbour’s holding) and the base which is equivalent to the phrase used in
§13, tir asa n-dssa.
cro ime thorad. DIL suggests that the O. Ir. form was a disyllabic croe, m.
/u-stem. We have kept the MS form because the evidence makes it doubtful
whether there was only one form of the word. Crau, O’Mulc. 208, seems old.
Moreover, in Mid. Ir. the usual forms are cro and cru though crua also occurs.
The changes ua> 6 and ua > ii are, however, exceptional even in Mod. Ir. (e.g.
snodh for snuadh ; see DIL under 1 snuad; and cf. B. 6 Cuiv, The Irish of West
Muskerry, §291). The Mid. Ir. exx., therefore, suggest two distinct O. Ir. forms:
crau/ cro (cf. goo for gdu as early as Wb. 5a8) and * era ii/ croe. Crau/ cro would
be the exact cognate of Mid. W. creu « *krawo-\ see GPC under crau and
Pokorny ZCP xviii 71 ; and compare gdu: Mid. W. geu < *gawa. Grammar §69,
gen. sg. gue <C *gdwias, Mid. W. noe <C *nawid, but O. Ir. nau <7 *ndwa).
The cro in question is presumably some kind of fence, but it is not easy to
see where it was supposed to be put. The phrase ime thorad suggests that it was
around the fruit in question, namely that which fell onto the neighbour’s land,
perhaps separating it from the rest of the neighbour’s land in order to delimit

io4
NOTES
[§12
the area into which the owner of the tree could go to recover his share of the fruit.
Alternatively, and this would make better sense, it was again around the fruit in
question, but separating it from the fruit deposited on land belonging to the owner
of the tree. The cro would then distinguish the fruit belonging solely to the owner
of the tree from that which he had to share with his neighbour. One can imagine
that, without a clear boundary fence, disputes could easily arise.
We read ime with the B gloss against im(m)a AC. This may be a case of
archaic unstressed e; see Introd. p. 13.
fo bith snuaid (?) tire. The interpretation of this phrase is rendered difficult
by the obscurity of the term snuaid or snuad. O’Davoren’s explanation (./. ecosc
no sainfeth) agrees with the explanations of the first and third glossators in A,
assuming that seanaith is a misreading of sainfeth. B’s gloss implies a quite
different meaning for snuaid. The possibilities will here be discussed in order:
(a) the reading, (b) the explanation given by the glossators in A and in O’Davoren,
(c) the explanation given by B, (d) other explanations in glossaries.
(a) The reading. A and O’Davoren both give a neutral final -d as opposed
to B’s palatal - d , and B has snaid in the lemma but rossnuaidter in the gloss. It
is, therefore, safe to assume that the reading reflected by B’s gloss was snuaid.
(b) The explanation given by the glossators in A and by O’Davoren assumes
that snuad is the reasonably common word meaning ‘ appearance, colour,
aspect ’, (1 snuad in DIL ). There are no early exx. which would determine the
inflexion of snuad. In Irish Grammatical Tracts (Declension §95) it is treated as
a m. «-stem which would require reading snuado here if that were the earlier
inflexion. The earlier exx. (apart from this one) are in chevilles, which suggests
only that a favourable sense was often given to the word. If, therefore, we accept
this explanation snuad may well mean something like ‘ good condition \
(c) The explanation in B apparently gives an otherwise unattested sense to
snua{i)d, something like ‘ damage ’ or ‘ trespass ’ since the purpose of the cro is
arna rossnuaidter a tir oc cuinchid ( sic leg.) in toraid ‘ so that the holding be not
damaged/trespassed upon while seeking the fruit.’ The negative shows that B
cannot be thinking of snuad ‘ appearance ’. The verb snuaidid is unattested (Mod.
Ir. snuadhaim ‘ I flow ’, Dinneen, cannot be relevant) and, therefore, gives us no
clue. It looks as though B is supposing that a fence should be put up to separate
the neighbouring holding from the one to which the tree belongs. This would
then separate the fruit deposited upon the neighbour’s holding from the rest of
the fruit and prevent the owner of the tree trespassing onto the neighbouring
holding, and perhaps causing damage, while collecting the fruit. It would also
prevent either party from collecting fruit to which they were not entitled. If,
then, we assume some such meaning as ‘ trespass ’ or ‘ damage ’ for snuaid, B’s
explanation makes sense.
(d) A number of other apparent homonyms of snuad ‘ appearance ’ are given
by glossaries : see DIL under 2 snuad — <5 snuad. None of them fit the context here
except, conceivably, 6 snuad ‘ herb, plant ’, but this is only a glossary word.
The explanation in B seems the best hope. The verb snuaidid in B’s gloss
cannot be derived from snuad ‘ appearance ’ since the latter seems to have a
favourable sense unless further qualified, and hence the negative in arna
rossnuaidter would be inexplicable. It seems likely that there was a word snuad,
of different inflexion from snuad ‘ appearance ’, that was still current when the
B glossator was working (probably ninth century) but went out of use soon
afterwards. B may just be guessing; the verb snuaidid may be only his own
creation; but it is more likely that his explanation is a genuine attempt and not
just word-play. He does not indulge elsewhere in such empty interpretation. In

§§12, 14]
NOTES
105
the absence of corroboration the explanation is of course uncertain, but provisionally we read snuaid, gen. sg. of snucid, an o-stem, to be distinguished from
snuad, //-stem, ‘ appearance and we assign it the meanings ‘ trespass, damage
inid-chuirethar a torad. This phrase occurs twice in §12. In its second
occurrence A has incuirither, but the inf. pron. can be restored from the gloss
in-ada-cuirither ( ada ‘ lawful, proper ’ is a frequent ‘ etymological ’ gloss on -id-).
We take the subject of inid-chuirethar in both instances to be the tree, and regard
-id- as a proleptic use of the inf. pron. {Grammar §421), anticipating torad
‘ fruit
In §13 the text has inid-chuirethar, not followed by a torad. We take -id- here
to refer back to §12 torad, and translate ‘ into which it (the tree) deposits it (the
fruit)’. Our reason for taking the implied subject of inid-chuirethar in §13 to be
the tree is the parallel between the two phrases tir inid-chuirethar and tir asa
n-assa. In the latter phrase ‘ land from which it (the tree) grows ’ there is no
doubt as to the implied subject: it can only be the tree. If this is admitted it
then becomes easier to take the implied subject of inid-chuirethar also to be the
tree. Otherwise there would be an awkward change of subject between the two
parallel phrases. It then becomes simpler to assume the same subject, the tree,
for inid-chuirethar in §12.
Dr. Binchy, on the other hand, would prefer to take the subject of inidchuirethar to be a torad in both paragraphs. He would regard the inf. pron. as
reflexive rather than proleptic, and translate ‘ into which the fruit projects
itself.’ This makes equally good sense.
cacha cethramthae bliadnae. We follow B’s reading cacha cethram- bl-,
taking it to be genitive of time {Grammar §250.4). One might also emend A’s
in cethramad bliadain to in cethramad mbliadnai, taking it to be accusative of
time (§249.3), or to ind chethramaid b/iadnai (or bliadain), taking it to be dative
of time (§251 .3).
§14
da llino. The //-stem lin fluctuates between masc. and neut. in O. Ir. {Grammar
§309). Here, the reading da Hina of A indicates neuter declension, as da is
followed by lenition when masculine, and by nasalisation when neuter {Grammar
§385).
imid-chomba. We tentatively emend iniclicoimge A, indacomgne C to imidchomba, 3 sg. pres. subj. of * im(m)-combuing [*imb-com-bong-] ‘ destroys ’
with 3 sg. neuter infixed pronoun, Class C (referring back to §12 barr-bunad).
The preverb and infixed pronoun are a relatively minor problem. The
emendation of in- to im- is justified by the gloss emh-cuimhges in the hand of
the first A glossator. E{i)m is regularly used as an ‘ etymological ’ gloss on the
preverb im(m)- (for exx. see AL vi s.v. eim; DIL s.v. 2 eim). Consequently, when
the main A glosses were composed (they are mostly older than the MS — see
Introd. p. 15) it is probable that the text had imid-. This seems to have been
corrupted during the course of transmission — perhaps by the main scribe — to
inidi- with ni for m and transposition of i and d. There may also have been
contamination from inid-chuirethar of the preceding §§12 and 13. We take the
form indacomgne C to be an example of the spread of the inf. pron. -da- to the
3 sg. masc. neut. This is already happening in the 3 sg. masc. Class C in O. Ir.

io6
NOTES
e.g. ruda-nordan ‘ which has dignified him ’ Wb. II 33c5. In SR -da- {-do-) is
attested a few times as 3 sg. inf. pron. e.g. 1. 3705 roda-car ‘ she loved him
One can compare §12 of BB where original inid-chuirethar appears in O’Dav. as
indocuirither , with -id- replaced by -do-. Also comparable is §32, where the
original probably had conid-tubart {comidubart A). The readings conatab-t E,
conatabairt F may go back to an intermediate form condatabart, with -id- replaced
by -da-.
Much more uncertain is our emendation of -coimge (glossed cuimhges) A,
-comgne C to chomba which we take to be from *com-bogs. The only other
clear example of the prototonic 3 sg. pres. subj. of con-boing which we have
noted is CG 519 arna ’ mmaconbba doib ‘ lest they destroy one another ’ (with
imma- as reflexive preverb). Here Dr. Binchy emends to -comba. DIL (s.v. conboing) also gives -coimb as a prototonic 3 sg. pres. subj. of this cpd. (with
reduction of the root, cf. Grammar §§626, 627). However, this form is very
doubtful, as its sole occurrence is in the rather corrupt A recension of Audacht
Morainn (spelled -choinb). Even there, it is the reading of only one of the three
MS-groups. See Kelly: Audacht Morainn p. 63 1. 90 v. 11., cf. p. 8 1. 56 v. 11.
We suggest that -coimge A is a scribal modernisation reflecting the
development of O. Ir. con-boing to Mid. Ir. coimgid. The scribe — or succession
of scribes — who changed -comba to -coimge may have been influenced by forms
with syncope, such as the plural past participle comgithi O’ Mule. 298 (from
*combgithi) where the cluster -mbg- would have been simplified to -mg- already
in O. Ir.
There may be a similar case in the O. Ir. Rule of Mo Chuta. In §41 of this
poem {Arch, iii 316) the MSS have rothoibge A, arcontoibge L, {dobera N). As
-toibge would not be possible in O. Ir. the original may have had some such form
as -torbais, 2 sg. perfective pres. subj. of do-boing (for a discussion of this stanza,
see forthcoming ed. by Prof. 6 Cuiv).
di-renar. This form occurs also in §§15, 35, 52, 53 and there are two instances
of the active di-ren (§§50, 54). The preverb di- is normally do- in the O. Ir. of
the Glosses {Grammar §83 IB) but di- is also quite frequent, so its presence in
our text cannot be used as dating evidence.
a besaib fidbreth. This reference suggests that there was a tract entitled
fidbretha ‘ tree-judgments ’ which is now lost. H.3.17 p. 658b {C1H 2106.3)
quotes what may have been the opening sentence of this tract: Can rosuidigthea
fidbretha la- ? ‘ whence have tree-judgments been established in Irish Law ?’
One can compare the final sentence of BB : is for sund rosuidigthea bechbretha la
Feniu. CU and BFG close with the same formula — see note to §55. There is also
reference to fidbretha in the Senchas Mar tract on distraint {CIH 388.18 = AL i
182) and in a late list of the legal topics which should be familiar to every judge
{CIH 2103.14 - ZCP xviii 363 §30).
The gloss ba [f]isgnce (in origin an ‘ etymological ’ gloss on besgne, e.g. CIH
408.2; 520.3) here glosses bes, cf. CIH 210.3; 468.10.
fo soiri each fedo, ‘ depending on the status of each tree.’ This must be a
reference to the division of the 28 commonest kinds of tree and shrub into four
classes, in accordance with their economic importance. The most valuable are
the seven airig fedo ‘ nobles of the wood ", next are the seven aithig fedo ‘ commoners of the wood ’, then the seven fodla fedo ‘ lower divisions of the wood ’ and
lastly the seven losa fedo ‘ bushes of the wood ’ {CIH 78.18 ff. = AL iv 146-8).
For suggested identifications see Celtica xi 107-24.

NOTES
IOy
§15
tairdbe. Here tairdbe must refer to the cutting of branches, as it contrasts
with bun-epe ‘ base-cutting ’ (§16).
coceniuil. The gloss is comaith cenel A raises the possibility of emending to
comcheniuil, cl. Stud. E. Ir. Law p. 46 §22 (= CIH 519.29-30) Mad be cetmunterasa
techta comaith 1 comceniuil — seek is comceniul each comaith , etc. Here comceniuil
is glossed ./'. commaith a cine I .
However, the composition form of com before c is usually con in O. Ir.
(' Grammar §830) e.g. cocad ‘ war ’ ( com + cath), cocele ‘ companion ’ ( com +
cede). Moreover, the form cocenel occurs at Sg. 198a 19, whereas comchenel is
not attested in the O. Ir. of the Glosses.
acht ni soeras nemed de, ‘ except that which nemed increases of it.’ The fine
for lopping a nemed tree is greater than for lopping a non -nemed tree. This
distinction is implicit also in Do fastud Chirt-\ Dligid {CIH 239.11 = AL v 474.6).
Here a list of offences entailing the same penalty include epe cacha feada acht
fidnemead no de{g)fidh ‘ the cutting of any tree, except a nemed tree or a sacred
tree ’.
We take de 4 of it ’ (glossed de sin A) to refer to dire ‘ fine ’ which is implicit
after di-renar, cf. §14 di-re nar lethdire.
nemed. The word nemed is used in a wide range of situations, many of which
occur in BB.
(1) Abstract. In the phrase acht ni soeras nemed de we take nemed to be an
abstract noun ‘ sacredness, privilege ’ with much the same meaning as nemthes.
That it is a neuter ostem is clear from instances such as neimed n-eclasa ....
neimed n-espuic ‘ the privilege of the Church .... the privilege of a bishop ’ in
CIH 588.1 (= AL iv 362). Other examples of the abstract use of nemed are
Romiodhair Neire nemhedh gach baird ‘ Neire adjudged the privilege of every
poet ’ {Eriu xiii 43.6 = CIH 1131.40) and Don-eim dano huasalneimeth righ no
espuic no cilli no primduine rig no ollaman filed ‘ the noble privilege of a king or
of a bishop or of a church or of the chief fort of a king or of a chief poet protects
him ’ {Biirgschaft §45 = CIH 594.30).
The Welsh cognate nvfet ‘ sanctity, privilege ’ is feminine (< *nemeta,
LHEB p. 279).
(2) Place. It is sometimes not possible to distinguish nemed ‘ sacred place,
sanctuary ’ from the abstract nemed ‘ privilege ’ above. Like (1), nemed (2) seems
to be a neuter o-stem cf. Gaulish nemeton, drunemeton, British Medionemeton,
Vernemeton, etc. (see Holder: Alt-celtischer Sprachschatz s.v. nemeton). In Sg.
13b nemed glosses sacellum ‘(small) sacred place ’ and in CG 54 it is used of a
sanctuary where a man cannot be pursued by his enemies: atnaig tar fot crunch
i n-ardnemed dia dlte, di-eim ar diantolae sliiaig ‘ he brings him away over gory sod
into a high sanctuary for his protection, which protects [him] from the swift
tide of a [hostile] throng ’ (for discussion of this passage see Binchy, Eriu xii
83-4, cf. B Crolige §21).
In the laws it seems to be used more commonly of any land owned by a
nemed-person (see (3) below) or by the Church. Thus CU {Eriu xvii 70 §11) says
that water cannot be conducted over nemed cille (church land) no duin (land
owned by a king or noble) no maigen fe[i]rt (a pagan( ?) grave-mound). According

io8
NOTES
[§15
to BG ( ZCP xxv 221) m gatar a uasalnemiud di-renar fo choibni trebe ‘ anything
which is stolen from a noble nemed- place is paid for as if it were a household
object.’
(3) Person. In BB the commonest meaning of nemed is ‘ a privileged person
that is, somebody who is privileged because of his rank or profession. It is a
masc. o-stem, e.g. §39 ind nemid (nom. pi.). The tract on status UB distinguishes
between soernemed ‘ free nemed ’ and doernemed ‘ base nemed ’ (cf. Audacht
Morainn §52). The soernemed- persons are ecalsi, fiat ha, filid, feine ‘ clerics, kings,
poets, nobles ’ ( CIH 1593.5-6 = AL v 14.10) and the doernemed- persons are
saeir, gobaind, umaide, cerda, legi, breitemain, druid ~\ aes caca dana oicena
‘ carpenters, smiths, coppersmiths, goldsmiths, doctors, judges, druids, and
people of every other art ’ ( CIH 1612.6-9 = AL v 90.19-21).
The glosses and commentary of BB (e.g. 42d, 44a, Appendix 1 (c), App. 2. (e))
distinguish two categories of nemed- person: uasalnemed ‘noble dignitary’
and iselnemed ‘ lower dignitary.’ There is no mention of the iselnemed in any
O. Ir. law-text, and it seems to be merely a grandiose term for ‘ commoner ’
employed in the later law-schools. This is supported by the substitution of grad
feine for iselnemed in commentary in H.3.17 (App. 3 (b)).
In the text itself there appears to be no distinction in meaning between nemed,
uasalnemed, and ard{d)nemed. Thus the uasalnemed ‘ noble dignitary ’ in whose
tree bees have settled (§36) is referred to simply as nemed in the next paragraph.
The same category of person is referred to as arddnemed ‘ high dignitary ’ in §42.
It seems probable therefore that nemed in our text always refers to the more
exalted members of the privileged class (kings, nobles, hospitallers, higher grades
of poet and cleric) and so corresponds to the soernemed of UB.
In the Heptad quoted in §39 there is a distinction (repeated in §40) between
nemed De ‘ a person privileged from God ’ i.e. a cleric and [nemed] duini ‘ a
person privileged from man ’ i.e. a king, poet, etc.
Finally, it seems that a person could be temporarily nemed and so immune from
legal processes, e.g. CIH 21 47. 1 0 nemed duine guba ‘ a person in mourning is nemed. ’
(4) Animal. In §§52-3 BB distinguishes uasalneimthiu {-nemeth B) cethrae
and lu-chethrai. According to the B glossator uasalnemid c[h]ethrae refers to
large animals (he mentions milch cows and trained oxen) while lu-chethrae refers
to lambs, piglets, calves and kids (for discussion see notes to §52).
The term nemid c[h]ethrae is used elsewhere in the laws of livestock which are
temporarily or permanently immune from distraint, cf. nemed duine guba in (3)
above. These include bo co ngalur ‘ a sick cow ’ nuithlech iar toud ‘ a milch cow
after calving ’, bo ciss flatha no eclasa ‘ a cow for the rent of the lord or church ’
{CIH 38.19 (= AL v 260); 897.11). Other nemid included in CIH 897 are oxen
when under yoke, bulls when the cows are in heat, race-horses at the time of
racing and swine being fattened.
When used of an animal (male or female) nemed seems always to be a masc.
0-stem.
(5) Tree. The most common meaning of the cpd. fid-nemed seems to be ‘ a
tree which grows on the land of a nemed- person.’ As the A glossator to 15c
points out, such a tree ( fidhneimhidh ) derives its sacredness from the rank of the
person who owns it rather than from its species. In the archaic legal poem
edited by Binchy {Celtica ix 157.40) the penalty for cutting a non -nemed tree of
the highest class {airig fedo) is one cow, whereas for cutting a fidnemed the penalty
is three cows (though this seems to be contradicted by H.5.15 26b {CIH 2183)
dire fidnemid ndir bo i s/an sluind ‘ the penalty of a noble fidnemed, declare a cow
for compensation ’).

§§15, 16]
NOTES
109
It is unclear whether cutting a tree of the less valued classes {aithig fedo and
fodlafedo ) would also entail a fine of three cows if it was growing on the land of
a nemed- person. The stiffer penalties attached to the cutting of nemed- trees reflect
the privileged position of the nemed-c\asses in early Irish society, and can be
parallelled from many other areas of law (e.g. Celtica x 34).
Fidnemed also has the more specialised meaning of a particular tree held in
veneration by the inhabitants of an area. The cult of sacred trees was very
important among the Celts (Anne Ross: Pagan Celtic Britain (London, 1967)
pp. 33-8) and survived well into Christian times in Ireland. Thus the word
de-fid (degfid MS), apparently used as a synonym of fidnemed in CIH 239.11, is
a compound of dia ‘ god ’ + fid ‘ tree ’ ( DIL s.v. defid). The Annals of Tigernach
(RC xvii 350.12) record the burning of a sacred tree (fidhnemidh) at Armagh in
995 A.D. (another word for such a sacred tree is bile : see DIL).
In Irish fidnemed (usually spelt fid ne(f)mid in later MSS) always refers to a
single tree, except in translations from the classics where it is used for nemns
‘ sacred grove ’. Among the Britons and Continental Celts there seems to have
been more emphasis on sacred groves than on individual sacred trees. Thus
Lucan refers to the Druidical rites which took place in groves ( Pharsalia I 450-8)
and Tacitus speaks of ‘ the groves consecrated to savage cults ’ in Anglesea
(Annales xiv 30). According to Strabo, the holy place of the 6th century Galatians
was called Drunemeton ‘ the sacred oak wood ’ ( Geographia xii 5.1), and in the
11th century Cartulary of the Abbey of Quimperle in Brittany, there is a reference
to silva quae vocatur nemet.
gloss 15c A. We are unable to make sense of ita in aind side in this gloss, and
suggest that the glossator omitted a word such as dire ‘ penalty ’ after in.
The spelling aind for and ‘ there, in it, for it ’ shows sandhi before the palatal
initial for side, cf. 17c, 49g aind sin. The spelling aind side is also found in a gloss
by the first glossator on BG (H.2.15A 38b7 = CIH 478.23). He glosses acht
ni etirscara lu de by ./'. it air is leitheineclann bias aind side ‘ i.e. since it is half of
the honour-price which will be for it ’. Hull (ZCP xxv 220) takes aind to be a ind
‘ its smallest amount ’ but this would give a *ta se fear sentence.
gloss 15c B. Dr. Binchy points out that the term etarbe lit. ‘ cutting between ’
seems to be used for the boundary between two holdings, cf. draignech bis i
fal etarba ‘ a blackthorn which is in a boundary wall ’ (AL iv 150.10 = CIH
582.20 etc.). The B glossator may therefore be correct in his claim that the cutting
of a tree in the boundaries of a man’s holding (i netarbib) entails a greater penalty
than cutting a tree in a wood.
§16
In the Laws, the seriousness of an offence sometimes depends on the season
in which it is committed. Thus the penalty for grazing-trespass is more severe
in winter than in summer (CIH 69.19 = AL iv 88), and the penalty for stealing
a billhook is doubled during ‘ the period of fencing * (CG 247).
This paragraph does not say what time of year lopping is regarded as an
equally serious offence as trunk-cutting, nor why. The B glossator suggests that
it is in summer and autumn, and is because of the fruit which is on the branches
(he also adds the unlikely suggestion that it is because of the swarms which may
alight during those seasons). The main difficulty with this explanation is that
it applies only to fruit-bearing trees, whereas the text has tairdbe each fedo ‘ the
lopping of every tree.’ The A glossator suggests that tairdbe (which he glosses

no
NOTES
[§§16-18
airdtimchell ‘ high cutting-around ’) entails two-thirds the penalty of bun-epe
(which he glosses lantimchell ‘ full cutting-around ’) during a month of growth,
but only one third during a month of dormancy. Although this explanation
diverges slightly from the text (where there is no mention of a two-thirds penalty)
it seems in general more probable than that of B. Less harm is likely to be done
to a tree if its branches are cut during the dormant months. If branches are cut
during the growth-period, disease may set in.
In the section of B Comaithchesa which deals with the cutting of trees ( CIH
78.15 = AL iv 146) there are some differences from BB. Firstly, the penalties
do not vary at different seasons. Secondly, it distinguishes three types of cutting:
base-cutting ( bun-beim ), the cutting of a fork ( gabul ) and the cutting of a branch
(i craeb ). BB, on the other hand, distinguishes only base-cutting {bun-epe) and
tairdbe , which we translate ‘ lopping ’ and take to include both fork- and branchcutting.
inid. The reading inad ‘ in which is ’ of B is the lectio difficilior and seems to
make better sense than as A. We restore O. Ir. inid.
comdiri. We read comdiri with B taking it to be predicative gen. sg. of the
noun comdire ‘ equal dire ’ cf. §51 it comdiri {-e MS) fri seotu trebe and Eriu xii
6 §3 comdire (read -/') a ben fri each recht. It does not seem necessary to distinguish
an adj. comdire ‘ of equal dire { DIL s.v.), cf. comgraid (gen. sg. of comgrad )
briugu friflaith ‘a hospitaller is of equal rank with a lord' (C/Fn608.8 = AL v 76).
§17
This paragraph deals with the division of the honey from a tree which grows
at or near the boundary between two properties, cf. §12.
in tire. For our emendation of itire A to in tire, cf. §44 acht is in tire i
suidigethar is coir bunad a tuisten. The emendation is supported also by the
definite article in the corresponding glosses: ind fearaind asa nfasand se A, in
tir asa nasaid na bich B.
in toraid. Our emendation of torad A to gen. sg. toraid is supported by the
gloss a thoraidh A. B has the ambiguous tor- in both text and gloss.
§18
Fintiu griain i mbechbrethaib. This phrase serves as a heading announcing
a return to the subject of bechbretha after the excursus on fidbretha, §§12-17, and
in particular a return to the topic raised in §§10 and 11, cobdaili saithe. The
legal metaphor mentioned in §§10 and 11 is here taken up and applied (cf. §11
hi fintedaib griain). The point of the metaphor may be seen from such a passage
as the following {CIH 2011.30 f. etc. Text normalized and emended):
Con-beir cinaid each brathair; cot-mbeir derbfine. Saigthech each fine,
erentach each fine iar n-elud chintaig. Dina-be derbfine, fo-loing iarfine a
chin *] a flaith, a rig, a recht. ‘ Every kinsman jointly bears liability for an
offence; the derbfine jointly bears it. Every kinsman/kindred is suable,
every kinsman/kindred is liable to pay as a result of the default of the
offender. If there be no derbfine, the iarfine sustains his offence (i.e. pays for
it) and [sustains his obligations to] his lord, his king [and his obligations
under] his ordinance.’

§§18, 19]
NOTES
III
Just as kinsmen sustain ( fo-loing ) the offence {cin) of a fellow-kinsman, so do
neighbours sustain {fo-loing) the grazing of the bees (§23), a grazing which can
be described as a cin, ‘ offence ’ (tin a mbel, Appendix 6 (b)). Moreover, the
notion of proximity of kinship, of close kinsmen, facilitates the analogy. Just
as the nearest kinsmen, those of the derbfine, sustain a fellow-kinsman’s offence,
leaving remoter kin only responsibilities contingent upon there being no close
kinsmen, so the nearest holding to the bees, tir bes-cla nesam, has sustained most
grazing and so takes the first swarm. The other holdings follow in order of
proximity to the bees.
This passage is notable evidence for the earlier role of the derbfine as the
primary kindred, the iarfine and indfine being supplementary. The primary role
was later taken over by the gelfine, as in the tract later called D' fodlaib cineoil
tuaithi, CIH 429.14 ff. (see also Binchy CLP 101 f. = Proc. Brit. Acad, xxix 222 f.).
This latter tract was probably compiled in the early eighth century, up to two
generations later than BB. In this paragraph in BB the derbfine is clearly primary
and the gelfine is not mentioned. CU also shows that the derbfine was the primary
kindred (see Erin xvii 59-60). Since BB and CU are the work of the same compiler,
it is highly probable that mention of the gelfine in §§11 is due to a later scribe,
though the B gloss llf suggests that the insertion was made during the O. Ir.
period. There were four cricha and hence such a scribe may be forgiven for
supposing that his text should have four fini: gelfine , derbfine , iarfine and indfine
(cf. glosses 18c’d cleirbine ar geiline).
In §18, however, the derbfine still has the primary role. Here the scribe has
made no change, the derbfine is identified with the land closest to the hive; and
it is the derbfine that takes th e fintiu griain, ‘ the kindred-share of land \ When a
man died without descendants the derbfine took and shared out his fintiu. The
fintin must, therefore, correspond to the cetsaithe of §19. The metaphor is
suggested by verbal parallels ( fo-loing , cin, nessam ); but its purpose is to set
out an order of priority between the claims of the neighbours. It justifies this
order of priority by appealing to an analogy with a more fundamental set of
rules in Irish law, those governing the inheritance of land. In the second note to
§2 it was pointed out that the text’s use of this metaphor involved serious
confusion.
The gloss on fintiu griain in B betrays complete misunderstanding. Because
B does not grasp the metaphor, he attempts to make fintiu griain into the place
where the hive is situated, but this makes nonsense of the whole passage: it is
the first swarm and not any land that the nearest holding acquires.
§19
The three years of freedom are now finished and the neighbours claim their
deolaithe. The neuter ed on refers back to tir bes-da nesam in §18; cf. CU §1 :
. . . tir asa tairidnither. Ar is ed on ailes each cetld ndeolaid seek na cricha o/chenae.
The fanciful interpretation of deolaid in glossb also appears in gloss7 to §1
of CU.
cetsaithe. The first swarm is referred to in a fragment of H. 3. 18 commentary
{CIH 663.17): Cetsaithi a ndire na saileach, i is as-sin gabar cona samaisc is fiu
in saithi do reir dligid hi cinn .ix. noidche iarna cur amach fic ‘ The first swarm in
payment for the willow, and it is from that that it is deduced that the swarm
is worth a heifer according to the law nine nights after its emergence, etc.’

1 12
NOTES
[§§19-21
The commentator has assumed (1) that cutting a willow entails a heifer as
a fine, (2) that a first swarm can be paid for cutting a willow, and infers (3) that
the first swarm (after nine nights) is worth a heifer.
§20
bes tech. Cf. as tech 4 which is best ’, Ml. 37d3, 73al0 and Grammar §139.
For the omission of a in bes tech torad, see note to §33 batar beich.
tanaise. For B’s gloss tarbsaithi see note to §25 glossa.
§21
Mat comoicsi. D1L under ocus gives a w-stem inflexion for both the noun
and the adjective; but under com(f)ocus it gives the adjectival inflexion as
o - a. The early forms are, however, the same: nom. pi. fem. oicsi Ml. 78c2,
comaicsi Ml. 26bll. If the adjective is inflected as an o - ci stem, the ending
in -i is in accordance with Grammar §353. Thurneysen, however, Grammar
§878, suggests that ocus 4 near ’ is related to Welsh agos, Corn, ogas 4 near ’.
Presumably he must think that the Welsh and Cornish words are borrowed from
Irish since Ir. g does not correspond with Welsh and Cornish g and it is difficult
to disassociate these words from Ir. oc 4 at, by ’ Welsh wnc, wng 4 close, near ’.
But, if this is correct, Ir. ocus should not have had an original front vowel in
the second syllable (cf. the o of Welsh agos), and, therefore, should not have
fallen under the rule stated in Grammar §353 (‘ Disyllabic adjectives with
palatal consonance resulting from syncope have the ending -i in the nom. acc.
pi.’). The palatal clusters in oicsi and comoicsi may, therefore, be secondary,
in which case the adjective may have w-stem inflexion. The noun comocus,
however, in the sense 4 kindred ’ or 4 kinsman ’ has o-stem inflexion: CA §33;
CIH 429.28; 430.5, 22; Stud. E. Ir. Law, pp. 140, 144, 148, 155.
commaith. The main scribe wrote coma, which was corrected to commaith
4 equally good ’ in a later hand, probably that of the fourth glossator. This
correction makes good sense cf. §20 bes tech (= dech ) torad.
fo-cerdar crann etarru. For the casting of lots when dividing the inheritance
of an extinct kindred, see CCF p. 42 (H Recension §65) and Thurneysen’s
commentary on Gubretha Caratniad §10 ( ZCP xv 317).
The casting of lots when distributing the swarms applies only to the second
and third of the surrounding lands. The compiler does not refer to the possibility
that all four may be equidistant from the hive.
da de. De here is an unstressed form of the numeral 4 two ’ as in cechtar-de,
nechtar-de see David Greene in Indo-European Numerals ed. Ross (forthcoming).
meraige. The third swarm is of less value than the first two, and is called the
meraige 4 fool * (see Introd. p. 47). In the gloss (21 c) and commentary (App. 1
(b), App. 4 (c)) the word has a prosthetic s: smeraighe.

§§22-24]
NOTES
113
§22
gloss3. The alternative explanation ‘ or at the end of that same year ’ conflicts
with the text. In the fourth year (after the three years of immunity) three swarms
are assigned to the first three neighbours. The fourth neighbour must wait
until May of the fifth year.
§23
dlegair. The subject is the clause ara mbe fer uaidib occa n-imchomet.
na rre sorche. This is correct observation: see Introduction p. 46. Re is
already fern, in most of the glosses, e.g. nom. sg. ind re Ml. 133b7, but is neuter
in the Carlsruhe glosses on Bede, Thes. ii 12.31; 18.33. Na ire could be either
(1) na neut. of nach ‘ any ’ + re acc. sg. neut. or (2) the article + re acc. pi.
neut. (either ‘ during any period ’ or ‘ during the periods ’). We opt for (1),
though the gloss na ree solusta could be taken as evidence for (2), cf. dat. pi.
/ n-inbuithib cuir saithe in §44.
Good early evidence for the gender of sorche is lacking. The antonym,
dorche, was a n. /o-stem (see DIL s.v.). In Mid. Ir. sorche is a f. id- stem, dorche
a m. /o-stem, but it seems likely that they were both neut. at an earlier period.
On the other hand, sorche may here be the adjective rather than the noun in
which case emendation to sorchi would be unnecessary provided that re were sg.
uaidib. DIL gives uaidib as a Mid. Ir. form and only ( h)uadib as O. Ir. This
is as a result of a less complete collection than that of Zeuss-Ebel: uaidib Wb.
24b29, 27dl, 3 1 dl 8 ; huaidib, 32a29.
arna erlat a saithi. As-lui is used of the swarm which gets away without
being tracked; cf. §§44-45. It should not be translated ‘ absconds ’ if bees are
in question, since that has been used as a technical term for the departure of all
the adult bees from a colony, and absconding is thus quite distinct from any
form of swarming: C. G. Butler, The World of the Honeybee , p. 13. In the year
in which the neighbours are to receive swarms they must provide a man to ensure
that the swarms are tracked.
§24
ro-ucca. We follow A and read ro-ucca, 3 sg. perfective pres. subj. of beirid.
The 3 sg. perf. ro-uc of D is also possible, but it is more likely that its omission
of -a is a scribal error caused by the following poss. pron. a.
a techtae. A’s gloss is correct as an example though not as a full explanation.
Techtae may refer to both the saithe received by each neighbour and to the
cain chuisc, allabrig n-ai and mian ngalair due during the three years of immunity.
All these form part of the techtae, ‘ due ’, of a neighbour.
ni dlig-aide tairgille na smachtu. Cf. note on §1. The implication that a
neighbour was entitled to either a tairgille or smachta, but not both, is apparently
supported by §26 and also by B Comaithchesa CIH 64.26-27 ; normalized :
Cach n-imbe cona smachtaib, each tairgille cona chaithchib, ar i mbiat
smachta ni biat caithchi, i mbiat caithchi ni biat smachta. ‘ Every fence
has its fines; every fore-pledge has its penalties; for where there are fines
there are no penalties, where there are penalties there are no fines ’.

NOTES
114
[§24
The caithig is the penalty for an offence against the land of a neighbour
through direct personal neglect or a deliberate personal act; for example, failure
of a herdsman to prevent trespass ( CIH 77.38-39: Ogcachach in sin ar is og a
follugud). A fore-pledge is given to guarantee payment. Fencing changes the
situation for then personal neglect has only indirect results. Failure to maintain
adequate fences leads to distraint on the offender or his kindred to compel them
to make good their neglect ( CIH 75.24f.) or to certain specified penalties: for
example, an axe is due for an oak-fence ( CIH 65.13, where they are termed
smachta, for no obvious reason). Trespass by animals in fenced pasture, where
no herdsman was required, is penalised by smachta ( CIH 66.32-71.15). But it is
not always clear that no tairgil/e was given to guarantee payment of smachta :
one may owe smachta tairgilli ( CIH 66.32), and the early poem edited by Dr.
Binchy, Celtica ix 157.23-5, assumes that a tairgi/le is given to guarantee payment
of smachta. But if the distinction between a caithig, in the sense of a penalty for
neglect or an offence committed by a person, and smacht, in the sense of a penalty
for trespass when personal neglect is not immediately responsible, is not always
maintained in the texts — which it is not, see CIH 69.5; 71.15 — then it is not
surprising that further distinctions concerning the need for a tairgille are not
maintained either.
The evidence of BB is not easy to reconcile with the exclusive alternatives
of B Comaithchesa (either tairgille and caithchi or smachta and no tairgille). BB
talks of smachta, § §7, 24, 26, 38 but never of caithchi, and yet according to the
statement in B Comaithchesa caithchi, not smachta, go with a tairgille.
One important issue is what it is that smacht refers to in BB. It is, after all,
a general term used in B Comaithchesa for a set of fines for trespass; the fines
mostly consist of sacks of malted barley ( CIH 66.35-36; 70.3-6). There is,
however, no suggestion in BB that its smachta consisted of such sacks. The
main clue to their nature lies in their relationship to the three years of immunity
enjoyed by a new bee-keeper. §7 shows that the immunity was from smachta ;
§24 suggests that, as soon as smachta have been paid once, the neighbours have
no right to demand other smachta or to retain the tairgille ; according to §26
(see notes) a man might be allowed to retain the tairgille and forego the smachta.
In §38 smachta are clearly related to trespass by bees and also to the period of
three years (the three years in which tor ad is shared coincide with the three years
of immunity for the new hive.) The smachta then are paid after the three years
of immunity but only once to any neighbour: once he has his smachta he has
no further claim. The fourth (and perhaps fifth) year is crucial for smachta : this
justifies the glossators’ conviction that smachta consist of one swarm to each
neighbour (e.g. §24c>d). Everything said of them suggests that they are the
swarms given gratis to the neighbours in the fourth and fifth years.
If we return to examine §§24 and 26 more closely they can be seen not to
offer any real support for the alternatives posed by B Comaithchesa. If a
neighbour has received his swarm from the bee-keeper, we can expect that
trespasses of the one hive of bees will now be cancelled out by the trespasses
of the other hive of bees. Neither tairgille nor smachta are now due: the tairgille
has performed its job of guaranteeing the giving of swarms to neighbours; the
smachta are excluded since the trespasses of the bees cancel out. Similarly in
§26 (see note) the neighbour who does not wish to receive a swarm may prefer
to take the tairgille. This has the advantage that, although he will probably
have no bees to cancel out the trespasses of his neighbours’ bees, the tairgille,
by passing into his possession, will no longer guarantee the payment of smachta.
Tie, too, therefore, will be excluded from making further claims against the
original bee-keeper. For him the tairgille can be an alternative to a swarm (i.e.

§§24, 25]
NOTES
115
the smachta ) because, first, the tairgille is presumably not very different in value
from a swarm, and, secondly, the tairgille for bees is returnable once one has
received one’s swarm. This, however, is not true of the tairgille of comaithches :
its function is permanently required while the neighbours continue to farm alongside their fellows. The choice of §26 (either tairgille or smachta) presupposes
conditions peculiar to the law of bees and has no relevance to the alternatives
posed by B Comaithchesa. In BB , then, a tairgille guarantees the payment of
smachta , just as it does in the poem edited by Dr. Binchy ( Celtica ix 157.25).
Even B Comaithchesa , as noted above, does not consistently follow its own rule
assigning the tairgille to caithchi rather than to smachta. This rule appears to
have been an attempt, made in a text compiled some time after the date of BB
— probably in the first half of the eighth century — to advocate a terminology
foreign to older texts.
ar is ed uathath insin con-oirg fri sochaidi la Feniu. This is a general legal
maxim quoted for the particular purpose of justifying a rule concerning bees,
i.e. the rule stated at the beginning of this §. It is also quoted in the commentary,
Appendix 1 (b). It should, then, be interpreted as resolving the possible dispute
between the neighbours who have received their techtae on the one hand and the
owner of the bees on the other. The neighbours are the many and the owner of
the bees is the one or the few. Provided they have received their techtae, they
have no further claim in respect of the bees: though they are in the majority
they now have no right to tairgille or smachta, and thus the one defeats the many.
This explanation is consistent with the B gloss but not with that in A. The B
gloss appears to add something to the text, and its early date is confirmed by
the neuter gender of lestar; it therefore deserves particular attention. The fer
tire is the neighbour who receives a swarm of bees as a deolaid and so acquires
a lestar. The bees from his hive will then graze together with the bees of the
other neighbours who have also received swarms and with the bees from the
original hive. Such joint grazing ( con-geil , congelt) implied a partnership in
which claims arising out of grazing-trespass would not normally arise; cf. CG
83 and note, con-gel{f)at ./. tir i mbi tuinidhi flatha la fine imalle, CIH 913.28-29
(O. Ir. gloss), and also the short tract (O. lr. but written as commentary) on
comingaire , ‘ joint-herding ’, CIH 576.24ff. As the main text does elsewhere, B
here uses the analogy of rules of comaithches, primarily concerned with cattle,
to justify a rule about bees.
§25
inscuchud. It is most probable that inscuchud here means nothing more than
a move whereby the hive came under new ownership by gift or, as the A gloss
would have it, by sale; cf. §27 cumscuchud. The B gloss raises the possibility
that the move in question is what bee-keepers term ‘ migrating ’ or ‘ absconding ’.
This is where a whole colony leaves its home due to lack of food in the locality
(Butler, The World of the Honeybee p. 154). But this is very unlikely to be the
intended meaning of our text: see following note on tola tarb.
gloss 25a. tola tarb. The B glossator, who probably worked in the 9th
century, explains mad inscuchud ‘ if [the colony of bees] should move ’ with the
words ni bi tola tarb aim doib do gres ‘ there is not an abundance of bulls for
them continually.’ We very tentatively suggest that tarb ‘ bull ’ is here used to
mean ‘ drone ’.
In an article entitled ‘ Bees in Indo-European languages ’ ( Bee World 55
(1974) 15-26, 46-52) D. E. Le Sage discusses the Indo-European words for

n6
NOTES
[ §25
‘ drone He points out that ‘ biological knowledge has brought us to dissociate
the male bee from that aspect of it which first made it conspicuous to our
forebears: its loud and low-pitched buzz.’ None of the words for ‘ drone ’ in
his collection reveals a knowledge of its reproductive function. It is therefore
unlikely that tarb refers to the maleness of the drone. Most words for ‘ drone ’
seem to be of onomatopeic origin, e.g. Sanskrit bambhara, Latin bombo, Spanish
zangano , Old English dr an ( >> Modern drone), Greek thronax. We suggest that
tarb is used for 4 drone ’ because the deep-throated lowing of a bull is similar
to the buzz of a drone.
But even if we are correct in our identification of tarb with 4 drone the B
gloss is still unclear. Why should the B glossator have thought that lack of drones
would cause a colony to move? One possibility is that he knew that migrating
swarms contain few or no drones, and assumed that this was the reason why
entire colonies sometimes abandon their nests. In fact, starving bees may
destroy their larvae and pupae — particularly those of the drones — sometime
before moving to another locality {The World of the Honeybee p. 154), so lack
of drones is a result rather than a cause of migrating. An even more remote
possibility is that tola tarb refers to a 4 drone assembly ’ where drones and queens
from a wide area congregate for mating (ibid. pp. 65-6). Both these explanations
presuppose a considerable knowledge of bees on the part of the B glossator or
his informant. However, it is clear from his glosse to §29 that he was sufficiently
expert to know that a bee dies if it leaves its sting in a person’s skin.
We have found no evidence that tarw ( tarv ) 4 bull ’ was ever used for 4 drone ’
in the British languages. The Welsh word used to gloss fucus 4 drone ’ is bygegyr
(see GPC and IEW 116). In Old Breton fucus is glossed satron vel guohi (the
correct meaning of guohi is 4 wasps ’ cf. Med. Welsh gwychi 4 wasps ’, Old Cornish
guhien gl. vespa). O. Bret, satron survives in the modern language as sardan
4 drones ’, sardanen 4 a drone ’ (Ternes: Breton de Vile de Groix, p. 203). In the
O. Corn. Vocabulary fucus is glossed by sudronen (read sadronen ). Apart from
the possible use of tarb in 25a, we have come across no Old or Middle Irish
words for 4 drone ’. In Mod. Ir. it is ladrann (lit. 4 robber ’ < Lat. latro) and
in Scottish Gaelic it is seillean diomhain (lit. 4 idle bee ’).
tarbsaithe. The word tarb occurs also in the B glossator’s term for the
second swarm (20b tanaise ./'. tarbsaithi lit. 4 bull swarm ’). A later commentator
uses the same term (Appendix 4 (c)). Through the kindness of Dr. Seamas 6
Cathain of the Department of Irish Folklore (University College Dublin) queries
were sent out to collectors in Cork, Kerry, Clare, Tipperary, Galway, Mayo and
Donegal to see if this term had survived in Modern Irish. Replies were received
from all these areas, but no trace was reported, apart from a possible use of the
term 4 bull swarm ’ in the English of Co. Tipperary.
Exact correspondances to tarbsaithe — both in form and meaning — are found
in Welsh and Breton. Llyfr Iorwerth (Appendix 7 p. 192) uses taruheit (tar6heit)
for the second swarm. The manuscript Peniarth 32 (our D) gives a variant
karuheit lit. 4 stag swarm ’ but this probably goes back to a mistranscription of
c for t in an earlier exemplar. In Llyfr Blegywryd the term used is eil heit 4 second
swarm ’, agreeing with secundum examen of the Latin Redactions ABDE.
Taruheit is obsolete in Mod. Welsh (Spurrell’s Welsh-English Dictionary s.v.
tarwhaid ) but the Breton cognate tarvhed is well attested from the modern
language e.g. Vallee: Grand Dictionnaire Frangais-Breton (1931) s.v. essaim, Le
Gonidec: Dictionnaire Breton-Frangais (1850) s.v. tarvhed, cf. Ernault: Glossaire
Moyen-Breton (1895) s.v. tarn. The agreement of O. Ir. tarbsaithe, Med. W.
taruheit and Mod. Bret, tarvhet indicates a Common Celtic *tarwo-satios 4 bull
swarm ’. A difficulty is the fact that W. heit 4 swarm ’ is fern, whereas its Ir.

§25]
NOTES
ii 7
cognate saithe is masc. However, as Bret, lied is also masc. it seems that Welsh
has innovated. See Watkins, Etudes celtiques xvi 191-4.
In both Irish and Welsh, the prefixing of tarb ( taru ) ‘ bull ’ to a noun generally
indicates fierceness, e.g. O. Ir. tarbflaith ‘ fierce ruler ’. However, as the second
swarm is not particularly aggressive, this explanation does not suit tarbsaithe.
In a letter, Ken Stevens of Bicton Agricultural College, Devon, has suggested
to us that the term ‘ bull swarm ’ may refer to the liveliness of the second swarm,
which generally contains a vigorous young queen. Mark Daniel of Crowthorne,
Berkshire writes: ‘ It is possible that “ bull swarm ” refers to the great number
of drones present in the second swarm, which usually has an unmated queen.’
He points out that the noise of the drones milling around the hive often becomes
a loud roar, audible 80 yards away. Perhaps, therefore, the term ‘ bull swarm ’
refers to the noise made by the second swarm.
It is very unlikely that tarbsaithe means ‘ mating swarm ’. In The World of
the Honeybee p. 64 Dr. Butler points out that a queen normally leaves on her
mating flight unaccompanied by drones or workers (see Introd. p. 46). The
mating behaviour of the honeybee is difficult to observe, and still not completely
clear (Butler pp. 63-7), so it is unlikely that the early Celts knew much about it.
Early Greek and Roman authors generally held that the honeybee did not mate
at all. Aristotle ( Historia Animalium v chapter 21) records the belief that bees
collected their young from certain types of flower. Virgil ( Georgies iv 295-314)
puts forward the theory that bees issued spontaneously from the corpse of a
bullock which had been clubbed to death. (For further discussion on early
views about the bee’s reproductive system, see note on Welsh modrydaf ‘ queenbee ’ in Appendix 7, p. 202.
ess. For the spelling ess for ass ‘ out of it, away ’ in this gloss, see Introd.
p. 9.
acht cain chuiscc. The lenition of cuiscc indicates that acht is here followed
by the nom. (see Index s.v. acht). Alternatively, it is possible that one should
emend to acht a ccain chuiscc , as in §6.
ar dlegair doib-sorn. The implied subject must be deolaid; cf. § §4, 9, and, for
the omission of deolaid , §10. The B gloss, .i. cobdaili saithe, is a correct
explanation.
soid fora techtae. For this idiom cf. Burgschaft §§29 and 32 = CIH 593.12, 17
(with do-soi ). For a different phrasing of the same rule that the rights of the
neighbours to gifts of swarms gratis become operative after the three years of
so ire, see §7. The idiom found here is to be distinguished from two other uses of
soid found in the laws: (1) soid X for Y, ‘ Y becomes liable to X ’ (e.g. CA §8 =
CIH 481.28-29: Mana airigt[h\er fair isin treide-seo ni sui meth for sin ceili de;
CA §43 = CIH 493.18: Ni sai for fine nach [n-]ecur n-inndligtech in-cuirither
[e]cond (u)ina(i) recht); (2) soid X fri Y, ‘ X turns to Y ’ (e.g. CIH 22.9-10: mac
fir rotuislither fo grad na tinntae i tuaith acht soid fri haithrige). In (1) the subject
is a word referring to some legal offence; in (2) the subject is personal. In the
construction found here and in Burgschaft §§29 and 32 the subject is a noun or
pronoun referring to the case in question, or else the verb is impersonal.
This paragraph is, therefore, simply an assertion that the general rule given
in §7 applies not only to new hives formed from swarms but to hives which have
moved and so changed ownership. A mere change of ownership with no change
of site could hardly confer three years of immunity, nor could a movement of a
hive without change of ownership, for then it would be a simple matter for the
owner of the bees to maintain perpetual immunity.

n8
NOTES
[ §26
§26
This is a very difficult paragraph, partly because it is difficult to be sure what
foma means, and partly because of the difficulties over tairgille and smachta
discussed in the notes to § §1 and 24. Recht is also a difficult term. A new subdivision of the tract begins with §27, the discussion of injuries to persons committed by bees. §26, therefore, should be a concluding note to the topic of
cobdaili saithe, divisions of swarms. Since §25 is a supplementary rule coming
after the main rule governing the divisions of swarms, §26 may well be another
supplementary rule. It envisages a preference, presumably on the part of a
neighbour, for a tairgille rather than smachta. The smachta in question must
be the swarms owed to the neighbours in the fourth and fifth years (cf. §7).
Hence it looks as though the neighbour in question in §26 is a man who does
not wish to receive a swarm as his techtae , his due. He does not wish to become
a bee-keeper. Yet the gift of a swarm deolaid, gratis , was to be compensation
for the grazing of his land by the bees. This, then, is probably the difficulty with
which §26 is concerned. It may be useful to continue the analysis of the first
clause (Cip dune . . . smachta) a little further before tackling the difficulties of
the rest of the sentence. The neighbour, according to this interpretation, may
have the alternative of a tairgille if he does not wish to receive a swarm. But up
to this point the tairgille has served to guarantee the gift of swarms after the
three years of immunity. In the ordinary course of events the tairgille would
have served its purpose once the neighbours had received their swarms. The
tairgille given to his neighbours by the man who first acquires bees is, in this
respect, quite different from the tairgille which is given according to the ordinary
rules of comaithches, for the latter guarantees the payment of caithchi, penalties
for personal offences or neglect, and thus is required as long as the parties
continue to be neighbours. The tairgille in ordinary comaithches is a permanent
requirement of neighbourhood; the tairgille of BB is attached to the cycle of
three years of immunity followed by the gift of swarms in the fourth and fifth
years. The tairgille of BB is thus temporary, unlike the tairgille of ordinary
comaithches. In §26 a neighbour is envisaged as preferring a tairgille at the point
at which it should normally have fulfilled its function, and should presumably
have been returned. The obvious interpretation is, then, that the neighbour does
not want to become a bee-keeper and would rather acquire permanent possession
of the tairgille. This would cease to be a guarantee of future smachta and would
become, instead, the property of the neighbour.
is foma do each recht oca mbiat. This clause remains obscure to us though
it is possible that the interpretation given by Cairbre in glossc III is correct.
First, different syntactical analyses are possible. One might read is foma do
each recht oca mbiat, in which case each recht would be the subject, or is foma
do chach recht oca mbiat, in which case recht alone would be the subject, or,
finally, one might retain the reading which we have given, following the glossators,
in which is is impersonal ‘it is Secondly, the meaning of foma is uncertain.
This example of the word (together with the quotation in O’Davoren) is the
only example given in DIL, which suggests reading fo ma or fo mu comparing
fo chen ‘ welcome Ma, however, is a word only found in Berla na Filed and
two of the three exx. in DIL are probably late. It is glossed maith and may well
have been extracted from an etymological gloss on maith itself. In any case, it
is difficult to see how this suggestion would yield satisfactory sense. Thirdly,
recht in the laws is a word of different and sometimes uncertain meanings:
(1) ‘ law ‘ authority ’, ‘ person of authority ’ (e.g. Irishes Recht p. 6, tract on

§§26, 27]
NOTES
119
dire, §5; B Crolige §21 ; Biirgschaft §61 ; CIH 400.31 f.), (2) ‘ condition of person,
class of person’ (e.g. Irisches Recht p. 64, tract on the fuidir, §4; CA §43;
B Crolige §§17, 43, 44; CIH 463.30). Thurneysen took recht in the second sense
to be a common spelling in the laws for richt ( ZCP xiv 380, n. on CA §43); but,
although the spelling richt does occur, it seems rather to be a special development
of recht ‘ law ’ etc. in the sense of ‘ legal condition ’ and then ‘ person belonging
to some legal class The best approach in all this uncertainty seems to be a
consideration of the meaning of the phrase recht oca mbiat. Ata X oc Y is used,
first, in a locative sense ‘ X is chez Y ’, and then to express possession ‘ X
belongs to Y In both cases Y should normally be a word for a person or
place. The further sense ‘ X is within Y’s power, or capacity ’ (e.g. TBC Rec. I,
ed. C. O’Rahilly, 2575) also requires a word denoting a person after oc. Where
the term governed by oc is not a verbal noun, these seem to be the three
possibilities in O. Ir. On this basis we should expect recht to refer to a person or
persons of some kind, since it cannot refer to a place. Moreover, if recht were
to mean ‘ law ’ here one would expect fo to be used rather than oc (Fel. Oen.
Jan. 1 ; Wb. 20al2). Cach recht in the sense of ‘ person of any class or condition ’
is particularly well attested in B Crolige (§§17 (two exx.), 43, 44), and it seems
that each recht in our text probably has the same meaning. This is the way the
glossators all understood the phrase. Cairbre’s gloss on foma, rogo ‘ choice ’,
will then give good sense. The implied subject in oca mbiat is likely to be beich,
as O’Davoren saw. The whole sentence will then mean that the decision as to
whether the neighbour will be allowed to retain the tairgille rather than receive
a swarm belongs to ‘ every class of person on whose land the bees are ’ or ‘ to
whom the bees belong ’. One can take this in two ways : the tract may be including
those neighbours who have accepted swarms as their techtae or he may only be
referring to the original owner of the bees. The first seems the more likely. If
the text were referring to the original owner only, it would have used a phrase
such as fer in bunaid (§38) rather than each recht oca mbiat. The latter seems
rather to be intended to include all those who have bees, both the original owner
and the recipients of the swarms.
§27
ruirset. We doubtfully insert an 5 in ruiret of the MS, and take it to be 3 pi.
pres. subj. perfective of rethid ‘ runs, rushes cf. §28 in-ruirset ( inruiret MS).
collud. When used with regard to bees, collud seems normally to mean
removing the honey, rather than destroying the hive or its occupants. In App.
5 (e) coillid refers to the legal removal of honey by the owner ( amser i coillter
in beich). In the present example collud may refer to the illegal removal of the
honey by a thief, or its legal removal by an agent of the owner.
During the extraction of the honey, particularly if it were done by a thief, it
would be common for some bees to be killed. However, the purpose of the
collud would have been to get the honey, not to destroy the bees. It is therefore
surprising that the 16th century A glossator, Cairbre, refers only to the destruction
of the bees {oca mar bad). A possible explanation is that the practice of smothering
the bees before extracting the honey may have become general by the 16th
century. See Introd. p. 43 and note to §41b.
cumscuchud. The glossator takes this to refer to the moving of bees from one
hive to another. The true meaning is no doubt wider, and would include any

120
NOTES
[§§27, 28
shaking or disturbance of the hive. A person who moves a hive may be acting
with malicious intent or with the consent of the owner. In either case the owner
is apparently free from liability for any injuries which may result.
gabal. The glossator believes that gabal refers to the taking of swarms
‘ from branches and from open places ’. We think, rather, that it refers to any
seizure of bees, whether in a hive, tree or swarm.
ind amsir. We emend in am to ind amsir (dat. sg. of a{i)mser ) on account of
the gloss isan re suthain. Re suthain is the normal ‘ etymological ’ gloss on
amser (- ir ) with re ‘ period ’ corresponding to am, and suthain ‘ long-lasting ’ to
sir (taken to be for sir), e.g. BB §4 iar n-aimseraib iarsan re suthain; CIH
213.30 teora aimsera .i. teora re suthaine. However, there is the possibility that
re suthain may also be used to gloss am, in which case our emendation would
be unnecessary. One can compare the occasional glossing of bes by ba-fis gnae,
the regular etymological gloss on besgne, e.g. §14c (see Introd. p. 18).
i tochumlat. We follow the glossator in taking i tochumlat to refer to the
emergence of a swarm of bees from the hive. The general experience of beekeepers is that bees are docile when swarming as they are gorged on honey
(Schofield: Teach Yourself Beekeeping p. 67). However, for some days before
the emergence of a swarm they may behave in a restless and inquisitive manner,
and are liable to become entangled in people’s hair, clothing, etc. Cases where
onlookers have been stung at the actual time of swarming have also been
recorded, and swarms have even been known to settle on the head or body of
an onlooker (who might be severely stung if he panicked). It should be stressed,
too, that a swarm contains up to 10,000 bees and emerges suddenly from the
hive with a loud and frightening noise.lt is therefore not surprising that swarming
bees acquired an undeserved reputation for aggressiveness. One can compare
Redaction E of the Latin texts of the Welsh Laws (ed. Emanuel p. 483) si autem
heit exiens hominem occiderit .... ‘ if an emerging swarm has killed a man
(Appendix 7 p. 198).
Alternatively, one might again regard the glossator’s explanation as being
too particular, and translate ‘ when they (the bees) are mobile ’. In this case
do-cumlai would refer to bees emerging from their hive at any time, not just for
swarming. However, if this is the meaning one might have expected some such
verb as do-imthiret ‘ goes to and fro ’, cf. timthirecht bech i llau anli ‘ the going
to and fro of bees on a fine day ’ TBDD 1169.
§28
in-ruirset. We emend inruiret to in-ruirset, 3 pi. pres. subj. perfective of
in-reith ‘ attacks ’. The gloss gibe foro roindsaiged seems to refer back to §27 cip
e forsa ruirset {ruiret MS), which we similarly take as 3 pi. pres. subj. perfective
of reithid. It must be admitted that indsaigid normally glosses saigid ( AL vi s. v.
saigim ) but it is hard to see how ruiret and inruiret of the MS could contain the
root saig-.
gloss 28a. Glossators and commentators regularly make a distinction between
four categories of person: (1) torbach co n-arm ‘an armed man on lawful
business ’, (2) torbach cen arm ‘ an unarmed person on lawful business ’, (3) esbach
co n-arm ‘ an armed man not on lawful business ’ and (4) esbach cen arm ‘ an

§§28-30]
NOTES
121
unarmed person not on lawful business This fourfold distinction does not
appear in the O. Ir. texts, and is probably a fabrication of the later law-schools.
It seems totally irrelevant in the present case.
na. MS nach for na 4 nor ’ is a Middle Irishism. See DJL s.v. 6 nach (nach).
annrecht. We retain the MS spelling, as non-lenition of / n r in consonant
groups is often indicated by gemination in O. Ir. e.g. diga/lre Wb. 1 8a 1 , annse
6d9 ( Grammar §120), cf. BB §1 annsam ( annsom A, ( a)ndsom CD).
The first glossator takes olcc to refer to the killing of somebody else’s bees
and annrecht to stealing them, but this distinction seems artificial. The third
glossator’s bualad na ces may refer to the striking of bee-hives with purely
destructive intent. More probably (as his gloss is an addition to I’s anndirgidetaidh
gaidi riii), it may refer to the thumping of a hive by a bee-robber with the purpose
of driving out the bees into an empty hive which he has placed nearby.
§29
is si-ede. We emend isesede to is si-ede as fuil is normally fern, in O. Ir.
(though there is evidence that it was originally neut., Erin xx 3).
saith. Our emendation of said to sdith is in accordance with the gloss.
gloss 29e. The B gloss refers to the fact — not mentioned by the author of
the text — that a bee normally dies soon after stinging a person, as it is unable
to remove its sting from human skin, and so ruptures the tip of its abdomen.
manis. The form manis in this gloss is irregular, as mani is normally followed
by conjunct and prototonic verbal forms. Thus mani with 3 sg. pres. ind. of the
copula is manid, and with 3 sg. pres. subj. is manip. Manis may be a scribal
error for manip or manid, influenced by is later in the sentence, or it may be an
otherwise unattested by-form of manid.
There is a possible case of 3 pi. manis 4 if they are not ’ in a short legal passage
on cows and calves ( C1H 1267.17-31 = 967.35-968.15). The MSS have os manes
carait comsa ( CIH 1267.24) and ocus mana sgara comamsa ( CIH 968.9), which
might be reconstructed as os manis carait chomsa 4 and if they are not kinsmen
in co-operation(?).’ But the whole passage is very obscure. See also Thurneysen’s
note on nis (= nidat) in ZCP xx 371 and Grammar §796, and on cis in Grammar
§457.
amal chach, 4 as in every other [case].’ This refers to the general rule that
the killing of an animal atones for any injury which it may have inflicted, cf. AL
iv 102.8 = CIH 192.24.
§30
In addition to A, the text of §§30-5 is preserved in E and F (for which see
Introduction pp. 10-11).
mad suil rochaecha. As the eye-closing reflex of the human is very rapid,
injury to the eye by a bee-sting is extremely rare. Dr. Eva Crane of the Bee
Research Association has kindly sent us a note on 4 Bee-stings and the Eyes ’

122
NOTES
[ §30
by Dr. C. Allan Birch in Bee Craft vol. 45 (1963) p. 17. He points out that the
sting of a worker honeybee penetrates for about 1 millimetre, whereas the average
thickness of both the upper and lower human eyelid is 3 millimetres at the free
border and double that at the base. Consequently, the sting cannot reach the
conjunctiva of a closed eye. There may however be damage from ‘ seepage ’
of venom.
Dr. Crane has also sent us two Russian references : an account of ‘ Eye
injuries caused by a bee-sting ’ in Oftalmologicheskii Zhurnal (1975) 30 (1) 55-6
and a description of two cases of men suffering bee-stings on the cornea in
Pchelovodstvo 36 (10): 54-5 (1959).
It would seem therefore that the story of Congal being blinded in one eye
by a bee-sting is unlikely, but not impossible. In later tradition {Fled Diiin na
nGed ed. Lehmann 1.315) Congal is said to have been stung when a baby,
whereas BB implies that the injury took place in adulthood (§32). We are
informed that bees are extremely unlikely to carry out unprovoked attacks on
young children, but are sometimes blamed for the misdeeds of wasps.
gloss 30d. The commentary on cin bech ‘ injury by bees ’ in CIH 316.37 =
AL iii 432 (the first words of which are quoted here by the second glossator)
remarks that ‘ the book tells of a hive [in recompense] for his blinding and
does not tell of two hives for killing.’ The book referred to by the commentator
presumably contained a portion of the Senchas Mar — including Bechbretha,
which indeed does not discuss the legal consequences of a person’s death from
a bee-sting or stings. Such deaths are very rare (Butler: The World of the Honeybee p. 137) but can occur when a person becomes hypersensitive to bee venom.
Death from bee-stings is mentioned in Recension E of the Latin texts of the
Welsh Laws (App. 7 p. 198) and, in closely similar words, in Anglo-Saxon,
Frankish and Spanish penitentials (Wasserschleben : Die Bussordnungen der
abendldndischen Kirche pp. 212, 316, 467, 534).
cocrann. A parallel is found in CIH 192.23 ff. = AL iv 102.6 ff. (which is
in O. Ir. though written as commentary). If an animal is killed in a jointly owned
herd and the culprit cannot be identified, a lot is cast on the whole herd; the
animal on which it falls is held to be responsible.
forsin lestrai n-uili. The sense requires the emendation of naile A to n-uili,
which is supported by idle of the A gloss and nil- E (though F has nail-).
The collective lestrae ‘ vessels, bee-hives ’ is formed from lestar ‘ vessel, etc.’
by addition of the suffix -id. This suffix is often used to form abstract nouns,
but its use as a collective suffix is not recorded in Thurneysen’s Grammar (§ §263-5)
or Pedersen’s VKG ii (pp. 16-7), though referred to by Stokes in his Criticism of
Dr. Atkinson's Glossary to AL i-v (London, 1903) pp. 47-8. Stokes’s examples
include cethrae ‘ cattle ’ from cethir ‘ quadruped ’, a ire ‘ satires ’ from der
‘ satire ’ and other more doubtful cases. Dr. Binchy suggests in Erin xvii 83
that fert(a)e may similarly be a collective in -id from fert ‘ grave-mound.’ Another
possible example is roilbe ‘ mountain range ’ which seems to be from ro- ‘ great ’
+ *slebe, a collective from sliab ‘ mountain.’ If we are right in taking lestar to
be a 5th-6th century borrowing from British (Introduction p. 44), the collective
suffix -id must have been productive at least until this period.
The suffix -io- forms a collective in a few cases, e.g. daire (orig. neut. ?) ‘ oak
grove ’ from dair ( daur ) ‘ an oak-tree ’, cenele (esp. in the phrase a cenele ndoine
‘ the human race ’) from cenel ‘ tribe, nation ’, aithre ‘ paternal kin ’ (declension
confirmed by rime, Eriu xxvi 84 § 18) from athair ‘father’, mdithre ‘ maternal
kin ’ from mathair ‘mother’.

§§30,31]
NOTES
123
dp lestar. We emend lestra A to lestar, supported by leastar of the A gloss
and lestar E (against leastr - F).
§31
ceta-rucad. E and F here give the better reading, as bericl brith is the normal
phrase for ‘ he passes judgement cf. §33 of BB and C1H 356.5. However,
cetaraced A (emending to ceta-rochet ‘ was first proclaimed ’) is just possible,
cf. citaroichet Ml. 44b4.
for Congail Chaech. Congal, like other names in -gal, appears to have
originally been an d-stem, except that the acc. sg. is -gal, though the dat. sg. is
-gall (O’Brien, Celtica x 224; Grammar §288). The phrase berid brith for seems
to occur with both acc. and dat.: berat brith fuiri Wb. 1 3al 1 ; berid brith for
are/e Wb. 29b9. Confusion between Conall and Congal is also attested in the
genealogical tradition, but it would have been facilitated by an exemplar which
had Congail which a later scribe might easily change to Conall (F) or Conaill
(A). Conall is a m. o-stem so that Conaill is not a possible reading; rather it
points to an original with Congail. It is curious that A and F should largely
agree here against E, while in general E and F are close. The changes made by
A and F must be quite independent. Genealogical MSS similarly confuse the
names Dungal and Dimlang.
Congal Caech is a person of some historical interest and the evidence of BB
is crucial. Here we shall confine ourselves to the evidence for the identity of
Congal Caech, and leave till later the problem posed by BB in giving him the
title ri Temro, king of Tara.
There is no doubt about his floruit. He appears in the annals from 626
until his death in 637 at the battle of Mag Rath. In the regnal list of the province
of Ulster he is given a reign of ten years and it is noted that he died i cath Roth
(. LL i 193). The date 639 for the battle of Mag Rath also has some annalistic
support but this discrepancy does not matter for present purposes (see J.
Bannerman, Studies in the History of Dal Riada, 6 n. 1).
There is little doubt that he belonged to the Cruithni, a federation of peoples
stretching from the lower valley of the Bann in the north to the present county
of Fouth in the south, and including, among others, the kingdom of Eilne in
the north between Bann and Bush, the kingdoms of Mag Fine and Fatharnae
(Larne) in the centre, and the kingdom of Mag Cobo in the south. Congal Caech
appears in the Dal nAraidi regnal list {LL i 195). Though the genealogists, as
we shall see, do not agree on his pedigree, they do agree in assigning him to the
Dal nAraidi. Since the latter became the dominant dynasty among the Cruithni,
this suggests at least that he belonged to the latter. For the distinction between
the Dal nAraidi and the Cruithni at this period, cf. Adomnan, VSC I 49 Cruthini
populi, but III 17 mocu Aridi: Byrne, Irish Kings and High-Kings, 108; O’Rahilly,
E1HM 344f.
There is considerable room for doubt, however, to which royal kindred
among the Cruithni Congal belonged. Not all genealogical texts include his
name in the dynasty of Dal nAraidi (it does not appear, for example, in H.2.7.
p. 155a and [he Book of Ui Maine f. 67r (old fol.), 15r (new fol.), col. a, 1.61 ff.);
those that include him do so in different places. In the early seventh century the
main line of the dynasty divided into two branches descended from two sons of
Baetan with the confusingly similar names Fiachrae and Fiachnae. In the midseventh century the descendants of Fiachrae (cenel Fiachrach, Patric. Texts
180.23-24), appear to have conquered the old kingdom of Eilne whose bishop
had his seat at Coleraine ( VSC I 50; cf. the name Dungal Eilni in their pedigree

124
NOTES
[§31
given below). The descendants of Fiachnae remained in control of Mag Line
in the south of County Antrim, and were usually the most powerful dynasty
among the Dal nAraidi.
The genealogists appear not to have been sure to which of these two dynasties
they should assign Congal Caech. Those who assign him to cenel Fiachrach
cannot agree on where precisely they should place him, as shown in the following
stemma :
Version A: Congal Caech assigned to cenel Fiachrach: O’Brien, CGH p. 326
= Rawl. B 502, f. 162b5; ZCP viii 336.1-3 (= Laud 610 f. 110b); ZCP xiv 113
(Lee. B = Book ofLecan f. 137v col. d.); Ballymote f. 168 of facs. col. a, 11.17 ff. ;
Book of Ui Maine , f. 72rb61 ff. (old fol.), f. 19r (new fol.):
Baetan
Fiachrae
ob. 608
Fiachnae
ob. 626
(Version B)
Becc
Scandlan/Scandal
ob. 646
Dungal Eilne (Dunlang)
ob. 681
Congal Caech
ob. 637
(Rawl., Ballymote)
Ailill
Cu-Chuaran
ob. 708
Fiachra Cosalaig
Congal Caech
(Laud)
Scandlan
Congal
(Lee. B)

§31]
NOTES
125
Since Congal Caech died in 637 Laud and Lecan B have plainly erred (unless
Lecan B s Congal m. Scandlain is a different person, which is unlikely) and
Rawl. and Ballymote have very probably affiliated him to the wrong man.
The chronology of Version B is not so difficult:
Version B: Congal Caech a descendant of Fiachnae Lurgan: Ballymote
(f. 168, col. e, 11.36 ff.), ZCP xiv 111 (Lee. B, f. 137r col. d, 1.20):
Baetan
Fiachnae Lurgan Fiachrae
ob. 626 (Version A)
Mongan Eochaid Iarlaithe Scandal/Scandlan
ob. 625 ob. 666
Maei-caich Conall Cummascach Congal
ob. 637
Chronologically Version B looks much more plausible than Version A, and it
has accordingly won favour with historians (Byrne, Irish Kings and High-Kings
287; Mac Niocaill, Ireland before the Vikings 156). Furthermore Lecan B has
the following ascending genealogy: Congal Claen m. Scandlain m. Fiachna
Find m. Baedain m. Echach m. Conla m. Caelbad m. Cruind ba drui {ZCP xiv
111).
A closer look at the evidence is, however, disquieting. The main evidence
for Congal comes in what one may call descending genealogies which enumerate
the sons of Scandal/Scandlan mac Fiachnai (or Fiachrach). We may examine
the following four witnesses: —
(1) Ballymote f.l68e: Tri meic Scannlan: Mael-caich, Congal, Cumascach.
(2) Lee. A (facs. f.l25v. col. b, 1.43): Tri meic Scandail meic Fiachrach .i.
Mael-caith, Conall, Cumascach.
(3) Lee. B (facs. f. 137r. col. d, 11.26 ff.) = Book of Ui Maine f. 70v col. b1
11.25ff. : Ceithri meic Scandail .i. Mael-caith {Ui Maine : m#c elcaig) 1
Conall -] Cumascach 1 Congal Caech qui cecidit a Muig Rath.
(4) verse in Lee. B and Ui Maine {ibid.):
Scandal dixit :
Ad-gen mo maccu uili
Mael-caith, Mael-ruain, Mael-ruibi.
Ad-gen Conall ar a dath,
-j ad-gen Cumascach;
1 ni aithgen mo bru
as mac Scandail Cu-citu.

126
NOTES
[§§31, 32
It looks very much as though a process of accretion has been at work,
favoured not only by the confusion between Fiachrae and Fiachnae (as 'in Lee.
A), but also by the presence of a Scandal and a Scandlan in the branches of the
Dal nAraidi dynasty, and by confusion between Congal and Conall (as in MSS
A and F of BB). The crucial question is whether Lee. A or Ballymote preserves
the original. There are three arguments for preferring Lee. A (apart from its
addition nieic Fiachrach ). First, apart from Ballymote , all the witnesses recognise
a son Conall, but Congal is recognised only by Ballymote and Lee. B, and in
the case of the latter only in the genealogy and not in the verse. Secondly, the
majority of the annals (AU (gloss), AT, AFM, LL) give Congal’s father as
Scandlan but Mael-caich’s father as Scandal; only CS among the annals gives
Scandal for both, although Fland Mainistrech, too, has Congal’s paternity as
mac Scannail ( LL iii 511). Perhaps Scandal and its derivative Scandlan were
interchangeable; but this seems unlikely, especially in face of the agreement of
the majority of the annals. Thirdly, it is evident from Version A that Congal
Caech, a figure famous in saga and probably the last of the Cruithnian kings to
be King of Tara, excited a certain interest among genealogists who were, however,
ignorant of his true lineage. The use of the epithet Claen, rather than Caech, in
the ascending genealogy in Lee. B suggests the influence of saga; moreover this
genealogy cannot be said to confirm the descending genealogy in the same text
for it is very probably extracted from it. The truth seems to be that there is no
good evidence for Congal Caech’s ancestry.
On the other hand, the efforts of the genealogists to insert Congal into the
ruling kindred of Dal nAraidi are symptomatic : if he had originally belonged to that
kindred, his pedigree would have been firmly fixed in the genealogical tradition.
Since Congal was almost certainly a Cruithnian, the likelihood is that he came
from another royal kindred of the Cruithni subsequently displaced by the Dal
nAraidi. The following have some claim to be early kingdoms of the Cruithni
(excluding those west of the Bann conquered in this period): Eilne, Boinrige
(Dal mBuain), Ui Derca Cein, Mag Line, Latharnae, Seimne and Mag Cobo.
Eilne we have already met: the name Dungal Eilni (ob. 681) suggests that it was
conquered by the Dal nAraidi at the latest in the third quarter of the seventh
century. That the old dynasty of Boinrige had been conquered by the late seventh
century is shown by Muirchu ( Patric . Texts 80.11-15); the appearance of a
Feidlimid Buan, a quo Dail mBuain , in a descending genealogy of Dal nAraidi
in Ballymote f. 168 is doubtless symptomatic. On the other hand, the way the
genealogists connect the kings of Mag Cobo and Latharnae to the Dal nAraidi
at a much earlier point in the genealogy (three sons of Crond ba drui, CGH
p. 155, f.l43a 52ff.) suggests that these dynasties were able to resist conquest
and thus had to be fitted into the ruling kindred through a legendary link. The
Ui Derca Cein are fitted in at the same point but by means of a second marriage
which indicates that this dynasty was, in the genealogist’s day, less closely allied
to Dal nAraidi than the others ( CGH ibid.). Congal, however, is more likely
to have come from one of the casualties than from one of the survivors among
the Cruithnian dynasties. Their defeat will explain why Congal’s ancestry was
so obscure to later generations.
caechsite beich. See first note to §30.
§32
Ba-ch ri Temro. On the connective -ch see Introduction p. 13. There is no
direct corroboration of BB's statement that Congal Caech was king of Tara, but
it remains plausible. It is easy to show why other texts might not include his

§32]
NOTES
127
name among the kings of Tara, and also why he could have acquired the necessary
political weight and prestige to gain the title. Later lists of kings of Tara (or
Ireland) were produced by men who accepted the theoryjLat the Ui Neill had
monopolised the kingship of Tara in the sixth and seventh centuries, whereas
the late seventh century list in Baile Chuind (ed. Murphy, Eriu xvi 145-51),
although plainly composed from an Ui Neill standpoint, did not recognise the
claims of several of their kings included in later lists. The evidence of Baile
~€huind is quite irreconcileable with the idea, embodied for example in AFM, of
an uninterrupted succession of Ui Neill kings of Tara (cf. Binchy, Celtic and
Anglo-Saxon Kingship 32-38; F. J. Byrne, The Rise of the Ui Neill and the HighKingship of Ireland 18). Two other Ulster kings of this period may have been
kings of Tara: Baetan mac Cairill of Dal Fiatach (Ulaid proper) and Fiachnae
mac Baetain of the Cruithni. Baetan’s claims appear in the genealogical tradition
( CGII p. 275, 156b 40f. ; p. 406, LL 330ab 55). Fiachnae may well, as Professor
Byrne has suggested {Irish Kings and High-Kings 112), be the Feachno included
in Baile Chuind. The battle of Moin Daire Lothair (563) was not the final blow
to the power of the Cruithni, but rather an early high-water mark of Ui Neill
conquest.
Admittedly, Copgal’s name does not appear in Baile Chuind. This is not,
however, a conclusive objection. Since Baile Chuind was composed from an
Ui Neill standpoint it cannot be appealed to as a final judge in these matters.
For the period of Congal’s kingship Baile Chuind has first Suibne Mend of Cenel
nEogain and then Domnall mac Aeda of Cenel Conaill. They are also given in
later lists (e.g. LL i 95-6). Suibne, however, was killed by Congal Caech in 628
and Domnall was probably not in a dominant position for some years in spite
of his defeat of Congal at the battle of Dun Ceithirn in 629. Of this battle
Adomnan says: In quo hello ut multi norunt populi Domnallus Aidi filius victor
sublimatus est ( VSC I 49). The phrase victor sublimatus est has been conjectured
to mean that ‘ Domnall Aid’s son became the king of Ireland after the battle ’
(Anderson’s note ad loc.). Yet it is notable that Adomnan, who was prepared to
term Diarmait mac Cerbaill totius Scotiae regnator (I 36), and who gives Domnall
a relatively prominent place in the Vita, only terms him rex valde famosus in a
prophecy attributed to Columba (I 10). Even though Domnall was termed rex
Hiberniae by A U at his death it is worth examining the real balance of power in
the years 627 — 637 more closely.
It will be useful to begin by considering the participants at the battle of Mag
Rath and the associated battle of Sailtire, fought on the same day. Domnall
mac Aeda’s allies were the sons of Aed Slaine (of Brega), the king of Mide,
Airmetach mac Conaill Guthbind {LL i 197, 1. 5923) and his son Faelchu {AT,
CS ). Congal’s allies were the Dal Riata (Adomnan, VSC III 5 : extract from the
Liber de virtutibus S. Columbae of Cummene, 657 x 669), and the Mugdorna
{LL i 197, 1. 5924). In the battle of Sailtire the victor was Domnall’s nephew
Conall Cael mac Maile-Coba who defeated the Cenel nEogain. Domnall,
therefore, depended upon the Southern Ui Neill, while among his enemies were
the Dal Riata, of Scotland and Ireland, the Mugdorna of S. Armagh and the
Cenel nEogain.
The hostility of the Cenel nEogain to Domnall is less remarkable than the
presence in his army of the kings both of Brega and of Mide. What makes
Congal Caech’s acquisition of the kingship of Tara plausible is, first, the stance
of the Cenel nEogain and, secondly, the bitter feud between the dynasties of
Brega ancjJVIide which can only have been healed, and then only in part, after
635. If CongaLCaech took the kingship of Tara, he did so through the feuds
of the Ui Neill.

128
NOTES
[§32
The beginnings of the feud between Brega and Mide are mentioned by
Adomnan (ESC I 14). Aed Slaine is there said to have been warned by Columba
that if he committed the sin of parricide (i.e. fingal , kinslaying) he would lose
the totius Evemiae regni monarchia held by his father Diarmait mac Cerbaill,
and would rule only over Brega (in gente tuo), and then only for a short time.
In AU we have s.a. 600, lugulatio Suibne me Colmaen Moer ... la hAedh Slane ;
and s.a. 604, lugulatio Aedho Slane . . . o Chonall mac Suibne. The course of the
feud is set out in the following stemma in which the names of those who died
in it are marked by iug. and the date: —
BREGA
MIDE
Cerball
Illand
Libr6n
Da Mac Libren
iug. 622
Conall Guthbind
iug. 635
Airmetach
iug. 637
E&elchu
ob. 637
Diarmait
Aed Slaine
iug. 604
Colman Mor
ob. 555/8
Congal
iug. 634
Ailill
iug. 634
Blathmac
ob. 665
Diarmait
ob. 665
Suibne
iug. 600
Conall Guthbind was responsible not only for Aed Slaine’s death but for those
of the two sons of Libren in 622 (which may even be a separate feud) and Congal
and Ailill in 634, before he was himself killed by their brother Diarmait. Some
kind of peace must have been patched up at that point, but it was, it seems,
incomplete for the regnal list of Mide (LL i 197) notes that Conall’s son Airmetach
was killed at the battle of Mag Rath by Lommainech, king of the Mugdorna
and foster-father of Diarmait mac Aeda Slaine. Presumably Lommainech was
moved in part by his friendship for Aed Slaine. Though Diarmait fought with
Domnall his foster-father may not have been so easily appeased.
Cenel nEogain also had good reason to oppose their distant kinsmen of
Cenel Conaill. In 569 Ainmere mac Setnai, St. Columba’s first cousin, was

§32]
NOTES
129
killed by Fergus mac Nelline, probably of Cenel nEogain. Ainmere was avenged
the next year by his son, Aed, DomnaH’s father, who then defeated and killed
Colcu of the Cenel nEogain in 580. Mael Cobo, Domnall’s brother, was killed
by Suibne Mend in 615; and the latter also defeated Domnall in 628. Cenel
nEogain was not, however, a united kindred: Suibne Mend belonged to the
Cenel Feradaig, and he appears to have had a co-ruler or rival, Mael-Fithrich,
who survived until 630, when he was killed by Suibne Mend’s brother Ernaine
mac Fiachnai. The genealogists’ and annalists’ view of their relationships is
set out below: the names of those who appear to have died through feud with
kinsmen are again marked by iug. :
(Cen61 Conaill) Niall Noigiallach (Cenel nEogain)
Conall
Fergus
Setna
Ainmere
iug. 569
(Cenel Feradaig)
Feradach
Fiachnae
Aed Suibne Mend Ernaine
mac Ainmerech ob. 628 iug. 636
Domnall Mael Cobo
ob. 642 iug. 615
Eogan
Muiredach
Muirchertach
mac Ercae
Nelline
Domnall
Fergus
iug, 570
Colcu
iug. 580
Aed
Mael-Fithrich
iug. 630
Conall Cael
ob. 654
It seems possible from the speed with which the killing of Mael-Fithrich by
Ernaine followed the death of Suibne Mend in battle with Congal Caech that
Mael-Fithrich may even have supported Congal in the battle, or at the least
have sat on the fence. If this is true it was the struggle between two branches of
Cenel nEogain which gave Congal his chance, For Tt' followed Suibne’s defeat
ojFDoranall -in J528 .
It also, however, gave Domnall a chance which, in the short run at least, he
exploited more vigorously than Congal. In the same year as his defeat by Suibne
and Suibne’s death at the hands of Congal, Domnall attacked the Laigin. With

130
NOTES
[ §32
this crech rig the annals mark the beginning of Domnall’s reign ( regnare incipit :
AU). Presumably they refer to the kingship of Tara, for Domnall was already
king of Cenel Conaill. In 629 Domnall defeated Congal at the battle of Dun
Ceithirnn, a few miles west of Coleraine. Adomnan’s reference to this battle
( VSC I 49) describes it as being fought between the Nellis nepotes and the
Cruthini populi, Ui Neill and Cruithentuatha. However, it is not clear that
Cenel nEogain participated one year before the death of Mael-Fithrich and one
year after that of Suibne Mend. Yet whether they did or not, the victory of a
king of Cenel Conaill here, in an area in which the Cenel nEogain had the more
obvious hopes of domination, is proof of Domnall’s power as against Cenel
nEogain (contrast AU 681).
There is no reason to suppose that the battle of Dun Ceithirnn was particularly
decisive, even though Adomnan thought it worthy of Columba’s prophetic
powers. The internal balance of power within the province of Ulster was in
flux and though in the event this was to be all to Congal’s good it was as yet
incomplete. He had only recently become over-king of the Cruithni, at the
earliest in 626, with the death of Fiachnae mac Baetain of Mag Line in battle
against Fiachnae mac Demmain of Dal Fiatach. Fiachnae mac Demmain,
however, was killed in battle against Dal Riata in 627; and it is from 627 tjiat
the regnal list dates Congal’s rule over the province of Ulster. In the same year
as”theT)atf Fe of Dun Ceithirnn, Mael-caich mac Scandail, rex Cruithne, defeated
and killed Connad Cerr king of Dal Riata, the victor of 627. Mael-caich was
probably king of Mag Line. Since we do not know his attitude to Congal’s rule
it is impossible to be sure whether this constituted a threat to Congal. He may
well have been acting in concert with Congal for Dal Riata was probably at this
stage still in the alliance with the Ui Neill, and Cenel Conaill in particular, to
which it had adhered since the previous century. It has been claimed by Bannerman ( Studies in the History of the Dal Riada p. 5) that this battle was primarily
a struggle among the Cruithni since among the Dal Riata casualties the annals
mention a Dicull mac Eachach, ri ceneoil Cruithne , who was probably king of
Latharnae ( CGH p. 155, 143b 1 f.). It is, however, unsafe to assume that Dicull
must have been fighting against Mael-caich: the annalist lists first among the
dead the kings and then two princes of Dal Riata. His list is not necessarily
organised according to allegiance. The only consistency to be found in the
political stance of DaLRiata at this point is opposition to the^urrently dominant
people in the province of Ulster: when it was Dal Fiatach they fought them and
won frioW it was the Cruithni and against them they lost. It is clear from the
extract from Cummene Find that the Dal Riata switched in the years after the
disastrous battle against Mael-caich to a new alliance with the Cruithni and
opposition to Domnall mac Aeda. This change will have immensely improved
the military strength of the ardri of the province of Ulster.
The immediate opportunity for Congal to gain the kingship of Tara was
probably created by a new phase in the feud between the dynasties of Brega and
Midg, In 634 Conall Guthbind slew two of the sons of Aed Slaine, Congal and
Ailill; in 635 he was himself slain by Diarmait mac Aeda Slaine. Conall Guthbind
had been victorious in battle against Leinster only in 633, so that if Domnall mac
Aeda was to sustain the position he gained in 628 by his crech rig against Leinster
he had to rely on the support of the king of Mide; and yet he could hardly
maintain himself as king of Tara against the active opposition of the sons of
Aed Slaine. Ionan tradition, sympathetic to the political fortunes of Domnall
mac Aeda and Cenel Conaill, the kindred of almost all its seventh century abbots,
was hostile to Aed Slaine. It seems likely that Domnall supported the king of
Mide. If, however, "the "sons of Aed Slaine did facilitate Congal Caech’s

§§32, 33]
NOTES
131
acquisition of the kingship of Tara in the early 630’s they would naturally have
changed sides once they had killed Conall Guthbind in 635, and would have
reverted to the normal Ui Neill opposition to the claims of any outsider to the
kingship of Tara.
We began with the bare statement of BB that Congal Caech was king of Tara.
If BB was indeed composed in the seventh century, its evidence must have great
weight. The only evidence to set against it, of equal antiquity, is the negative
evidence of Baile Chuind ; but that text, as has been observed, was not disinterested.
It is unlikely that BB was composed in a law school in Ulster, and hence its evidence
is the weightier. The probability is, then, that Congal Caech was rilTemro for a
year or two in the period 630-635. BB itself shows that he lost the position
before” the Battle of Mag RathT
conid-tubart assa ftaith. A’s comidubart may be explained as a misreading
of conidubart = conid-tubart. That the first element is the conjunction con is
supported by gloss13 gur cuiredh he. The graphic assimilation of the final d of
cotiid with the nazalised initial of tubart would have been easy, since in ordinary
speech the final [5] would have been assimilated to the following [d]. We take
the forms conatab-t E, conatabairt F to reflect the spread of the Class C inf.
pron. -da- from fern, to masc. (i.e. conid'>conda'>cona ); cf. Eriu i 174 for
examples.
The u attested by A’s -dubart, against -tab(air)t EF is early O. Ir. (see Introd.
P- 14).
E and F also innovate in substituting fiaithemnas for flaith ‘ kingship,
kingdom, lordship ’. This sense of flaith appears to have become less familiar
during the Mid. Ir. period, cf. Thes. ii 326.4 where gloss5 has flaithius and gloss6
explains flatho by .i. flaithemnasa Herenn.
Congal Caech’s deposition can be compared with other losses of sovereignty
in Early Irish tradition. The blinding of Cormac mac Airt in one eye is said to
have prevented his entering Tara (Expulsion of the Dessi; Y Cymmrodor xiv
104-6 §§2-3; LU 4377-9). It is not actually stated in any O. Ir. text that he lost
his kingship, but the whole tenor of the 9th c. Tecosca Cormaic implies that he
had abdicated in favour of his son Coirpre Lifechair. The late introduction to
Bretha Etgid is more explicit ( CIH 250.13-15 = AL iii 84.3-7): ba geis rig co
nainim do bith a temraigh , *] rocuired cormac amach da leighes co aicill araici
temair ... 1 tucad rigi nerenn do coirpri lifechair mac cormaic. ‘ it was tabu that
there should be a king with a blemish in Tara, and Cormac was put out to be
healed at Aicill near Tara . . . and the kingship of Ireland was given to Coirpre
Lifechair, son of Cormac ’.
For other examples see Eriu xii 148 §§9-10 ( Tochmarc Etaine ) and Eriu xvi 38
§§6-7 (The Saga of Fergus mac Leti). For an example from outside Ireland
(Sassanian Persia) see Procopius, History of the Wars I, 11, 4 falso blinding in
one eye).
§33
To-bert a chin forsin fer. The syntax and meaning of the sentence do-beir X
a chiniaid) for Y derive from two simpler types of sentence: (1) berid X a chinlaid )
(e.g. Burgschaft §76d; CIH 74.26; 2011.30) and (2) a chin for X (e.g. Cain
Lanamna §36 = Stud, in E. Ir. Law 74; CIH 2011. 13, 16, 21; similarly teit a
chin for X, e.g. CIH 2011.19; biid a chin for X, CIH 2011.24, 26). In type (1),
without a prepositional phrase for Y, the subject is the man who is liable for
an offence; in type (2), if a person is referred to in the prepositional phrase for X,

132
NOTES
[ §33
it is he who becomes liable for the offence. He is usually not the person originally
liable, whereas in type (1) the subject generally is the person originally liable.
In our case the owner of the bees becomes liable as a result of Congal’s action in
charging him with liability for the bee-sting and the subsequent judgement
confirming the liability. We have a clear distinction, therefore, between berid X
a chin and do-beir X a chin for Y.
A’s reading a chin is probably correct as against EF a cin : as the phrase cin
oenfir, §34 shows, it is thought of as the offence of one bee not of them all.
batar beich. Here E and F have the better reading as against A’s non-relative
batair. According to the rule given in the Grammar §507 one would expect
batar to be followed by the possessive pronoun a ‘ his ’. Thurneysen’s rules
govern genitival relative clauses which are copula sentences. Of the constructions
distinguished by him two are as follows:—
A. copula + predic. adj. + subj. In this type there is no possessive pronoun
expressing the genitival relation :
(1) don brathir as enirt menme, Wb. lOcl, and not
(2) *don brathir as enirt a menme, nor
(3) *don brathir as-a enirt menme (though this construction is found in
verse.) :
B. cop. + predic. noun: “ the possessive pronoun is always inserted between
the relative form and the substantive itself e.g.
(4) fir as-a c[h]athach, CIH 245.29 = 424.2, and not
(5) * fir as chathach.
But these rules may not be universally valid for legal prose :
A. (2) is exemplified in CIH 491.25; each mac beo[a]thar nabi saer a c[h]or
= 1794.4, gach mac beoathur nadbi saor a chor ; but 2052.34-5, each mac
beoathar nadbid saor cor, keeps to the pattern of A(l).
B. (5) is exemplified in CIH 193.30 do fir bede heich = 577.32 do fir bete
heich; CIH 193.34 fer bide heich but 577.36 fer beta heich, perhaps bete +
a, conforming to B(4), cf. din fir beta beich CIH 578.31.
In verse a parallel to fer batar beich is provided by Blathmac Poems ed. Carney
1. 166 in fer batar congbala. Moreover, although the use of a when the stressed
possessive pronoun ai is the predicate is compulsory according to Grammar
§507(e), a contrary case is to be found CIH 577.20: berid in fer bes ai mart a
aighe. The reading is confirmed by the other MS, CIH 192.29, so that though
it would be easy enough to emend to bes a ai to conform to Thurneysen’s rule,
it is doubtful whether one should. In this case the unstressed possessive a occurs
before aighe, not before ai as in Thurneysen’s rule. It is likely that Thurneysen’s
rules are only correct for standard O. Ir. and that a more flexible system obtained
in early O. Ir. as in ordinary O. Ir. verse. Whether a in asa etc. was originally the
possessive pronoun, as Thurneysen would have it, or derived from -yo, as Liam
Breatnach argues (Erin xxxi 1-9), is irrelevant to the issue here.
noch is si breth inso. Here noch is clearly just a connective, although DIE
gives it as the first ex. of an explanatory meaning to be found in the laws. We
accept Dr. Binchy’s argument ( Celtica v 87-8) that in the laws sech is and noch
is are in essence only connectives, as against the view expressed in DIE that
noch is is (a) contrastive or adversative in Wb. and Sg. and occasionally in Ml.,
(b) explanatory in Ml. and in the text of the laws. It is essential to distinguish
between two things: (1) the meaning of the connective itself, and (2) the semantic

§33]
NOTES
133
relationship between two clauses joined by the connective. For example, ocus
may introduce a clause which is merely added to the previous clause, or serves
to explain the previous clause, or is contrasted with the previous clause (see exx.
in DIL under ocus). But these are not separate meanings of ocus. The examples
of noch is in BB illustrate the same point :
(1) noch is ed ro-saig ind n-oircne (§8). This whole clause may be construed
as introducing an explanation, but noch is is not thereby explanatory.
(2) The ex. in this paragraph. Noch is introduces a clause which is neither
explanatory nor adversative.
(3) noch ni asu a beim in chrainn fo bith ind nemid 11a asu dano a tuaslucud
(§36). We translate ‘ for it is not easier . . .’ because noch ni introduces
an explanation, not because the phrase noch ni is itself explanatory in
meaning.
(4) noch as-rulai saithe (% 45). Plainly just connective.
(5) noch is si a cuit-side (§49). Noch is introduces the specification of the
amount of their shares.
These examples show that the meaning of noch is is compatible with different
semantic relationships between the clauses connected. The same is true of the
examples of noch is etc. in Wb. The collection of examples in Zeuss-Ebel 700-701
(fuller than that in DIL) makes this clear: Wb. 10b9 cannot be contrastive; in
Wb. 19c4 noch ba introduces an explanation. Moreover, one of the examples
which is cited in DIL is incorrectly interpreted, perhaps because of the wrong
translation in Thes., Wb. 5al8: .i. is follus isindi nachimrindarpai se noch dm
amisrahelde gl. Numquid repulit Deus populum suum ? Absit; nam et ego Israhelita
sum. Stokes and Strachan translate ‘ i.e. it is manifest in that He has not cast
me away, although indeed I am an Israelite ’. To absit corresponds is follus with
the negative verb ; isindi introduces the explanation why it is manifest ; nachimrindarpai se noch dm am israhelde is the explanation and corresponds to the
Latin nam et ego Israhelita sum. The translation ‘ although ’ obscures the
correspondence between noch dm am and nam et ego .... sum. It should be:
‘ i.e. it is clear in that he has not cast me away and I, indeed, am an Israelite ’.
God has not rejected the people of Israel, because he has not rejected me and I
am an Israelite. In Wb., then, noch is introduces clauses which are contrasted
with the preceding clause or the Latin text, but also clauses which are explanatory.
In itself it is simply connective.
la Ultu -) Feniu. The lawtracts and associated tales maintain that Ireland is
divided between three free cenela : Feni, Ulaid and GailnilGailedinlLaigin. These
three are the primchenela , the cenela soera. A late glossator, CIH 356.15-16
introduces the Erainn, but an earlier glossator, CIH 883.33, still preserves the
old division: .i. feini 7 ulaid -j laigin. That the Eoganachta are Feni is shown by
lawtracts of Munster origin ( Uraicecht Becc: CIH 2256. 13, 14; 2262.20; 2217.12;
2279.1 ; Corns Bretha Nemed : CIH 2211.3; Erin xiii 14.20 = CIH 1112.11; 17.1,
6, 25 = 1113.40; 1114.4,18; 19.13 = 1115.24.). Interest in relations between the
Feni and the Ulaid is characteristic of the Senchas Mar though it is by no means
absent from Munster material. The tract on athgabal opens with two ‘ leading
cases ’ both of which appeal to legendary events involving the Ulaid and the
Feni. The Feni and Ulaid are each given a legendary jurist, Brig (a woman) for
the Feni and Sencha (or in other places Sen) for the Ulaid (CIH 209. 12-13, 22-23 ;
380. 14ff.). The O. Ir. tale of Nin mac Magach (CIH 907.36-908.6), which relates
a leading case on tellach, turns on the same division between Feni and Ulaid.
To judge by early glosses in H.3.18 (CIH 908-909) the leading cases in the tract

134
NOTES
[§§33, 34
on tellach all concerned relations between the Ulaid and the Feni. In the
Privileges and Responsibilities of Poets , there is only one reference to the Gaileoin
{Erin xiii 20.18 = CIH 1116.12) but four references to the Ulaid. (It is doubtful
whether Laighne at 34.22 = CIH 1126.8 is a genuine example of the name
Laigin). In this respect the standpoint of the legal texts is the same as that of
the Ulster cycle except that they tend to see things more from the side of the
Feni (notable examples are the superior wisdom of Brig, CIH 209. 22-23 and the
threat, Mairg d’Ulltaibh madh ala Boinn beid, Erin xiii 20.4-5 = CIH 1116.1).
Admittedly, in the laws, Feni often refers to the free Irish in general so that
la Feniu means, in effect, ‘ in Irish law ’, and Fenechas is the traditional oral law
of the Irish; but there still remains a consciousness of heterogeneous origins^
With this belief may be contrasted the doctrine, developed in the O. Ir. period
by the genealogists and the synthetic historians, of the single origin of the
soerthuatha as opposed to the aithechthuatha (cf. O’Rahilly, EIHM 154ff.). The
older Idea of the tri chenela soera rannsaite in n-insi so ( CIH 356.6) suggests an
explanation of the origin of the term coiced ‘ province ’. The term coiced can be
used of a holding of land which comes to a man through sharing with his kinsmen
{Thes. ii 238.5,6, coiced Caichain; 238.16-17, cenel Caichain = Patric. Texts
172.14-15, 25-26). If this term were applied to the sharing of Ireland between
three free cenela , and still more to the sharing between the sons of Mil Espaine
in the synthetic doctrine, this would account for the origin of the term. There
is, perhaps, no need to search for five primitive provinces: the shares will be
those of the Ulaid, Feni and Laigin.
Our text, in restricting the judgement on the kingship of Tara to only two of
the three cenela . the Feni and the Ulaid, may have been influenced by the relative
decline of the power of the Laigin: in the fifth century, it seems, the Laigin had
been active contenders for the kingship of Tara and had held lands in Brega, but
this was no longer true in the seventh century (cf. Byrne, Irish Kings and High
Kings , 137-8). The Lflaid, however, still are serious contenders until, at least, 637.
§34
nad foruachtatar. The gloss in A, *i nochor fuactnaigedar etc. suggests that
we have to do with a form of fo-fich. We read foruachtatar, 3 pi. perf. of fo-fich,
taking foruchatatar in A to be an example of transposed letters. This is supported
by the readings forfuachtadar E, for(f)uachtadar F. For the use of the /-pret.
with this verb, see Grammar §682.
fo-reccar la cona. The MS readings and the A gloss show that the verb is
fo-ricc and not fo-fich : they give no support to the emendation fo-rechar. The
B gloss to this sentence is clear enough until .?. dolinat selba fri seilb fortuit mat
illselbu. This may be a quotation. It may be interpreted with the aid of the short
O. Ir. text CIH 577. 19fF. The relevant portion (normalized) is as follows:
It e a mbretha inso tra mani festar guinide forsa n-ag riam. Dia festar, berid
in fer bes ai mart a aige -] as-ren a cheile ag fo ’laill do; -] mad ed ro-orr a
n-ag, fo-certar crannchor forru a aurlunn selb dtis da dib dia-toth bibdu do
thabairt; ~\ da-mbeir iarum inti dia-tuit -| inerenad a cheile do a aurlunn selbi
cethrae. ‘ The preceding are the (appropriate) judgments on them if the ag
(bovine animal) be not known as a gorer previously. If it be known, the
man to whom belongs the carcase of his ag takes it and his partner pays him
a substitute ag; and if it be this one (the dead ag) which has gored the ag
(the killer), lots are cast for them according to the pattern of landholdings
to know on which of them it may fall to give the guilty one (to the other
partner); and he upon whom it falls then gives it and let his partner pay him
according to the pattern of landholdings and cattle.’

§34]
NOTES
135
Reference is made elsewhere to the casting of lots to determine the ownership
of land: Gubretha Caratniad §10 ( ZCP xv 316), CCF (H Rec.) §65. These both
refer to the sharing of land between kinsmen; but the analogy referred to in the
phrase a aurlunn selb must refer to the allocation of an indivisible asset (or
liability) between claimants of equal standing, or of an asset which the claimants
agree not to divide between them but to allocate by casting lots.
In B gloss e for-tuit appears to be a form of the prepositional relative for(s)atuit with elision, or omission, of the relative particle -(s)tf-. It therefore seems
that the quotation envisages a liability falling upon one selb among many selba.
This makes it likely that do-linat is 3 pi. pres. ind. of do-lin in the sense ‘ swarms,
abounds ’ rather than of do-llna fri ‘ adds to ’ which yields no clear sense.
nadid-lamethar airthech na fortach. Both text and meaning are uncertain:
cf. §§44 ro-ch-lamethar forgall etc., 45 nach laimethar nech luge bed lais, and inna
laimther forgall na airthech bid neich int sainreduch. The MS readings in A,
aurthuch na forthuch , forgall , luigu, forgull na airthiuch, suggest that one should
construe these terms for swearing as instrumental datives. Although A can
write u for a in final syllable it only does so occasionally. The regularity with
which u appears in these phrases makes it difficult to argue that the u is merely
scribal. The other MSS, however, do not confirm these readings: for this
paragraph we have the testimony of F, which has a rather than u (note that
fortach is written out in F but omitted in E). In §45, B has luga for A’s luigu. So
far as one can see there are three independent witnesses: A, B and EF. Readings
suggesting the dative are confined to one of these witnesses. It would, however,
be very unsafe to argue merely on the basis of the apparent concurrence of B
and EF : genuine forms with o- and /ostem datives in -u- and -iu are more likely
to have been removed than introduced in the course of MS transmission.
Provided that one can exclude the possibility that the dative forms are merely
scribal then they must be early readings. Some legal MSS have a penchant for
employing 11 where it is not justified (notably H.3 . 1 7), but A is not one of them.
We have then, so far as the MS evidence is concerned, two possibilities demanding
serious consideration, and we shall begin by examining the syntax of the phrases
given an interpretation following A, that is, accepting that airthiuch etc. are
indeed in the dative case.
The syntax of ro-laimethar is characterized by frequent ellipsis of the logical
object, namely a noun or nominal clause denoting the action someone dares to
perform or the role he dares to assume. The following examples may help to
explain the difficulty of interpreting the use of ro-laimethar in BB:
(1) ro-laimethar is transitive:
(a) the logical object is also the grammatical object:
Ni lomethar nech dul don imdae ( TBDD 1207),
‘ No one dares to go to the apartment ’.
(b) the logical object is not the grammatical object:
brethem na laimetar gell fri hiumcosnum a brethe ( CIH 1377.39 =
AL v 352),
‘ a judge who does not dare [to give] a pledge in respect of dispute
as to his judgment.’
Laumur ar dochondaib dilsi cailli ( Celtica ix 157.44-5),
‘ Let me venture for [the benefit of] the immature [to state] the
immune things of the forest.’
(2) ro-laimethar is intransitive:
ma ro loidib lammis {Fel. Oen. Feb. 27), ‘ if we dare in lays ’.

136
NOTkS
[ §34
In (1) (b) the deleted logical object is shown in the translations within square
brackets: [to give], [to state]. If the first example under (1) (b) were re-written
without ellipsis, it would read as follows:
brethem na laimet[h]ar tabairt ngill etc. (or gel l do thabairt ).
As a result of the deletion of the logical object (tabairt), what had previously
been an objective genitive dependent upon tabairt, namely ngill, becomes the
grammatical object of na faimethar. The intransitive use of ro-laimethar, as
in (2), may have the same origin. The example given above is perhaps elliptical
for ma ro-lammis a mmolad i lloidib, ‘ if we should dare [to praise them] in lays ’
(using normal prose word-order).
The examples in BB divide into two groups, the present example and the rest.
The latter do not contain infixed pronouns and thus pose a slightly easier problem.
We may take as an example §44, reading forgull with A :
ro-ch-lamethar forgull in-otat in saithe hi tir a cheli,
‘ and he ventures with an oath that the swarm enters into his
neighbour’s land ’.
The curious feature of this construction is that the logical object is not deleted
but transformed into an instrumental dative. Forgall is the logical object. If,
therefore, this sentence were to follow the pattern illustrated under (1) (b) above,
it should read as follows:
ro-ch-lamethar in-otat in saithe etc.
‘ and he ventures [to swear] that the swarm enters ’ etc.
It might be said in reply that forgall, fortach and airthech are not such general
terms that they could be omitted without ambiguity; but this objection does not
apply to luge or lugu in §45, and it will not explain why these terms should be
put into the dative rather than into the accusative.
If one follows B, E and F and reads airthech etc., then the syntax is entirely
straightforward. A direct parallel is now available in an extract with a fairly old
gloss (with the deponent) preserved in two places in H.3.18 (CIH 824.29; 1087.36).
We give a text on the basis of both witnesses: Con-fodlae riasiu ar-tois d. fodail i
nde in feich dlighther and mana lamhathar nech a airthech, ‘ Thou shouldst divide
jointly before thou swearest a vicarious oath; i.e. the debt that is owing in this
case is divided into two unless someone should venture a vicarious oath in respect
of it (Note, however, that the gloss will only explain the text plausibly if
riasiu, O. Ir. resiu, has its etymological meaning ‘ before this ’, in which case one
should translate: ‘ Thou shouldst divide jointly ; before this thou shouldst swear
a vicarious oath ’).
In §34 A’s nadhidlamiter appears to contain an infixed pronoun. Since
concord is not always preserved ( Grammar §421), the pronoun may refer to
cinaicl or even to fer. The nearest parallel may be TBC Rec. I 718-9, dosrenga a
fortgae *] a forgaimniu ro bdtar fo Choin Culaind uair ndch rolamair a duscad, ‘ he
tugged at the rugs and skin-coverings that were under Cu Chulaind because he
did not dare to awaken him ’ (cf. 468: ni laimthe-som do duscad ; we are indebted
to Professor E. G. Quin for these two examples). In 718-9 ndch rolamair may
contain an infixed pronoun if ndch is not merely Mid. Ir. for nad. The infixed
pronoun could be either masc. or neuter and duscad could be acc. or dat. (cf.
468). Assuming that ndch contains an infixed pronoun, the possible literal
translations are as follows:
(a) ‘ because he did not venture it, his awakening ’;
(b) ‘ because he did not venture him, /his awakening ’
\ with his awakening

§34]
NOTES
137
Only the last of these translations will provide a parallel to §34 of BB. It should
be noticed that Y and W have a different reading in 468: ni Iaimtis-[s]eom
diuscad, ‘ they used not to dare an awakening ’ (Y); this conforms to the usual
type, (1) (a) in the classification above.
The difficulty in accepting the readings of A is not only that they lack
corroboration from other MSS. They also lack clear parallels in other texts. If
they followed the pattern of (1) (b) above (< brethem na laimetar gell etc.) all would
be plain sailing, but they do not. We have preferred to read airthech, for gall etc.,
therefore, for three main reasons: A’s readings are not confirmed by another
MS; there is no certain parallel for what must be a most complex construction;
there is a direct parallel if one reads airthech.
Nadhidlamiter in A is also difficult textually. Although E is partially illegible,
it does not look as though it agreed with F’s reading; hence the latter is probably
a recent innovation in the MS tradition. Similarly, the apparently pass, form in
E, laimt-, receives no support from noch ni laimither F. It is difficult to decide
whether to emend A’s -lamiter to sg. -lamethar or pi. - lame tar : F’s reading and
the A gloss argue in favour of the sg. Following A’s nadhid-, we read nadid- (a
rare use of nad- with infixed pronoun, Grammar §419) rather than emending to
nachid-.
The meaning of the clause is, in general, clear (cf., as before, §§44 and 45).
The particular significance of airthech is not so evident: in §44 forgall is used
alone, in §45 (opening sentence) luge, and (closing sentence) forgall and airthech.
On fortach and forgall see CG legal glossary under aire forgill and luge ; for
airthech see Binchy, ‘ Irish Faw and Irish History, II ’, Studia Hibernica xvi 21.
In the text known as Mac Ardde's Synod or The West Munster Synod ( ZCP viii
315f.) a number of saints are said to have guaranteed on behalf of their tuatha
that the latter would abide by an eternal alliance, bithbrathirse, between them.
The patron saints gave this guarantee by means of an airthech : dodeochaid tra
aurthach Brendain *j Mochuta Rathin ar Chiarraigiu . . . 1 aurthach Comgdin
airchindich Imleochu Ibair dar cend forthuath, ‘ the vicarious oath of Brendan and
Mochuta Rathin went as surety for the Ciarraige . . . and the vicarious oath of
Comgan, superior of Emly, on behalf of the fortuatha ’. The term airthech can,
then, be used with the phrase do-tet (or teit ) ar ‘ goes surety for ’ : a vicarious oath
may guarantee future performance. Cf. ZCP viii 316.27-9: ni thiagat hi sluagheth
for cecli tuaith de Mume frisi ndernsat brathirse fobith aurthig na noeb dodeochaid
etorro *j rl. In Conall Core and the Corco Luigde ( Anecdota iii 59) Conall Core’s
mother-in-law, father-in-law and wife, each in turn, compel him to stay a further
year in Albu by a vicarious oath :
Trath atchondairc mathair a mna-som ani-sein gaid essomon bliadna do -]
aretoing. ‘ Ni rag thorat,’ ol se, ‘ em dia taithmenaid dia bliadna,’ ol Core.
Aretoin[g] in ri in blia dain aile. Aretoing a ben in tres blia dain. ‘ When the
mother of his wife saw that, she begged him for a year’s respite and swore a
vicarious oath for him. “ I shall not go past you (i.e. I shall not violate your
oath),” said he, “ if you call it to mind in a year’s time.” The king swore for
him the next year. His wife swore for him the third year ’.
Just as the patron saint can swear an oath on behalf of his client tuath and to
guarantee its behaviour, so can affines compel performance by a vicarious oath.
In TBC Rec. I, 1150, Foeg says: ‘ Artung-sa deu . . . firfassa ardchless ar belaib
eirred .’ Here he can scarcely be swearing an oath for the gods, rather he swears
in the presence of the gods (a different meaning of ar-). In the passage from
H.3.18 already quoted the vicarious oath appears to determine liability to debt:
the division into two may be upset by an oath which presumably must be sworn

138
NOTES
[§§34, 35
by a third party. It looks as though the precise responsibility was uncertain so
that a rough and ready division into two was the standard solution. This is much
the closest parallel to our passage: here uncertainty again requires a rough and
ready solution unless airthech or fortach are forthcoming. The available material
does not suffice to answer fully the two main questions: who was entitled to
swear a vicarious oath? and on what issues was such an oath admissible? Family
relationships are plainly relevant to the first question, the uncertainty of the issue
relevant to the second, but that is as far as one can go. In our passage it must
presumably be some third party, not under suspicion of injury or killing, who
swears on behalf of a person who might be liable. It is difficult to see how this
might work in the case of a bee-sting, but a third party might easily have seen
which animal had killed another or which man had been the slayer. Even then
the immediate companions of the slayer, because they were his companions, were
not allowed to clear themselves by oath (dithech), nor to fix the guilt on one of
their number by oath (fortach). The oath of the third party would then replace
an oath sworn by the participants in an act, and so be vicarious, and would fix
the guilt on one man. In the case of the animals the third party was presumably
someone who did not own any of the animals who might have been the killer.
His airthech would replace an oath by an owner.
Airthech is not attested in the O. Ir. Glosses. In later MSS it is very often
spelled with aur-, ur-, or er- (e.g. BB §34 aurthuch A, urtach E, urrthach F against
§45 airthiuch A). The preponderance of spellings with (a)ur- lead to the suggestion
that the cpd. is * air-fo-tong - (Pedersen, VKG ii 653; Marstrander, Lochlann ii
213). However, as ar-toing is well attested (see DIL s.v.) we take there to be only
one preverb before the verbal root, and read airthech. Thurneysen points out
(Grammar §823) that (a)ur- spread from cpds. with air-fo- and air-uss- to cpds.
with only air-. This development is rare in the O. Ir. Glosses, but very common
in later MSS, cf. BB §1 taurgillib A for O. Ir. tairgiliib (see Introd. p. 3).
§35
This rule brings to an end, and summarizes the upshot of, the whole section
on personal injuries by bees, §§27-35. In these paragraphs the following injuries
have been mentioned :
(1) §§28 and 29: a man is stung by bees although he has not provoked them;
(2) §§30-33: a man is blinded in one eye by a bee-sting;
(3) §34, first parallel: a dead animal is found among dogs or pigs or cattle;
(4) §34, second parallel: a man has been slain in the midst of a crowd.
There is a feature common to all these situations: it is usually impossible to
identify the precise animal or the man responsible. If the particular bee which
has stung a man has been killed by him, that compensates, according to §29, for
the bee’s offence. A similar principle, however, governs the cases in which the
identity of the man or the animal is unknown, only here the group to which the
man or animal belongs, and which itself can be identified, is held liable (§34
ar-tet sochaide cinaid n-oenfir). With bees the liable group is a collection of hives,
a lestrae (§30), in effect the owner of the bees (§33). But allowing for the substitution of a collective liability for the liability of an unidentifiable individual,
the same general principle applies: either compensation or vengeance. This

[§§35, 36
NOTES
139
principle is to be seen in §29: if the bee has been killed no honey is to be paid in
compensation. The phrase amal chdch shows that this is only the implication of
a more general rule. The application of the principle to the situation of identifiable group but unidentifiable individual is stated in §35. The application is
stated only in general terms, and only the compensations for injuries by bees have
been specified (§§29 and 30), and vengeance in the case of an ordinary sting. But
by inference and the evidence of other texts it is possible to fill out the details as
shown in the following table:
Offence Compensation
Vengeance
§29 sting
sufficiency of honey killing of bee or, if compensation refused, forfeiture
( dilse ) of bees.
§30 blinding in one eye
§34 dead animal
§34 dead man
hive chosen by lot from
group of hives.
equivalent animal
(perhaps chosen by lot)
honour-price and wergeld
paid
group of hives forfeit
herd forfeit
sluag liable to vendetta.
For the case of the dead animal cf. CIH 192.23-33 = 577.13-24. For the dead
man cf. CG, legal glossary, under enech and eraic. The B gloss uses the term
eraic, but probably not in its narrower technical sense.
do-rochratar. The common uses of the perfect (or present indie, or subj. with
ro) in BB are (a) to state the act which brings about a particular legal state of
affairs and (b) to state the establishment of a particular rule (e.g. §11 rosuidiged).
Perfects of type (a) alternate freely with presents (e.g. §44 oa n-elat, but §45
as-rulai). In this example dilse is a state of affairs brought about by the failure
to offer proper compensation and the perfect is properly used for what causes
dilse.
In the first B gloss inidrochreta (divided in id rochreta in CIH ) is obscure to us.
Cuicir may perhaps be a term for a kindred, cf. Celtica ix 157.21 and note;
Stud. E. Ir. Law pp. 144, 148-9.
§36
The following tables set out the various divisions of the torad given in §§36-49,
together with the ultimate ownership of the swarm. Facts which can be inferred
but are not stated are placed within square brackets. The tables demonstrate
how important it was to track the bees: unless a tracked swarm settled in a crcinn
n-uasalnemid, the original owner was always entitled to recover possession. If
he had failed to track them, yet was able to swear that they originated from his
hive, he could only claim a share of their torad for three years. The owner of the
land where they settled could claim possession. Even when he loses possession
the original owner has the advantage that he can claim a share of the torad for
three years. When he recovers possession, he only has to allow others a share
of the torad for one year. Finally it is almost certain that when a swarm of bees
is found, but not tracked nor sworn to, the owner of the land acquires possession
while the finder gets a share of the torad. Only when, as in §49, there is no owner
of the land is the swarm and its torad the absolute property of the finder.

140
NOTES
[ §36
2 cc
00 c
O ctf
-5 £
£ C/i
"S3
g
O
'hi
4-1
o
WO
<D
Vh
o3
-C
wo
o3
g
[op
"EU
O
o
I 8
o
' WO
rC V- 1
00 3
"S ®
c
<2 . T3
I o §
O
> . C/5
G H
«i £
£
<D O
>>
o*
00
c
"Eh
o3
-G
wo
wo
D <D
4h !_ ’Zi
Out!
sf 8
s
> T3 ,u
°l s
GO
GO
<D <D
pC <u
£ «>
g £
03 c3
’E £
C/5
43
O.
c3
Vh
00
o3
Vh
o3
a
&
Vi
z
£
O
hJ
<
Z
HH
o
HH
cd
o
><
ffi
P
w
u
<
cC
Ec/5
LU
W
ffl
’TD
C
cd
Vh
<D
C
£
o
Vh
<D
C
>
o3
G
[op
"Eh
O
h|co
H|w
m
"S3
CD
r«*
s:
cd
'53
wo
53
"53
CD
s
<D
53
CD
<3
CD Si
•SP o
53 ^
-S §
t: -5
<3
rm
1
vo
cn
Vh
<D
c
£
o
13
c
[op
*Vh
o
«h*
H-'f
“53
CD
r—
S3
CD
«>*
I
c
o
G
CN
H
<
Oh
Q
OL
h- *
a
H
Z
CQ
Q
W
u
<
c<
H
C/5
w
w
PQ
Vh
CD
G
£
o
13
c
• ^
op
"Eh
o
Vh
CD
44 ,
o IH|CC
a
Vh
"S3
CD
5
cd
1
G
O
c
<43
•S?
53
r*.
S3
m
r t ■
3S
Co
£
s
■<£ CD
■> -s:
> .a
Q I
uj ^
^ .Su -s
c
c<
H
z
D
<4i
S3
£
O
*43
G3
C
-03
V-H
O
Vh
CD
C
£
o
m
"S3
cv
s
CD
c
o
G
53
S3
0*,
*43
"S3
<D
2
Wj
Si
■J3
Wo
V.
<D
o3
G
GO
5h
00
CD
G H<m
O
O
13
PC 5
< >2
^ §
in ^
s -
w -S
W o
U R
*< £"
OC O
H -2
Z
3 ^
CD
4-
*4:
VJ
C3
CD
^3
CS
>2
Wo
53
On
XS
G
03
<D
G
O
m
~cs
<D
I
G
O
c
in

§§36, 37]
NOTES
141
shares of tor ad
paragraph
where found ?
when?
finder
owner of
land
46
faithche thechtae
within 1 year of
swarm settling
1
4
3
4
47
faithche thechtae
after 1 year
1
2
1
2
48
sechtar faithchi co
rricci ruud mar etc.
—
1
3
0
jSl
3
49
ruud etc.
—
1
—
uasalnemed. In BB there appears to be no distinction in meaning between
the terms nemed, uasalnemed and arddnemed. See notes to §15 nemed (3).
noch ni asu a beim in chrainn. We emend nack of the MS to noch, translating
‘ for it is not easier to cut the tree ’ (for noch see notes to §33). This explanation
gets some support from the B gloss ar ni hurusa a imdibe where ar seems to indicate
an original noch in the text. This gloss occurs out of order in B (at the end of
the gloss on §17, see CIH 923.40). It was presumably taken from §36 on account
of the fact that both §§17 and 36 deal with swarms which settle on trees.
We take the a of a beim in chrainn to be the possessive pronoun, lit. ‘ its
cutting of the tree ’ i.e. ‘ cutting the tree.’ It could also be the neuter definite
article, in which case it would be necessary to emend to a mbeim.
§§36-8 deal with bees which have occupied a cavity in a neighbour’s tree,
whereas §§39-42 deal with bees which have (temporarily) swarmed on a branch.
The latter can easily be removed by the bee-keeper without damaging the tree,
but to remove the former would entail severe tree-cutting, and a consequent
heavy fine if this was done without authorisation. It is likely that under these
circumstances the bees were usually left in situ, and the bee-keeper had to be
content with one third of the honey for three years.
Unauthorised cutting of trees for the purpose of extracting bees and/or
honey is mentioned also in Welsh law (e.g. Appendix 7 : Redaction D 9C).
na asu. Note omission of copula, cf. Grammar §865.
glosse. Note the insertion of the conjugated preposition as between tuaslugud
and its dependent gen. na mbeach.
§37
do-etet. We follow Thurneysen ( Grammar §§627, 771) in taking do-etet to
be the same verb as do-autat (see DIL s.v.) which he analysed as *to-ad-uss-tet
(. ZCP xii 410) against Pedersen’s *to-in-tet ( VGK ii 644). For the alternation
-1 au -, -e-, -/-, in this verb see Grammar §80 (c), e.g. dan-autat, LU 6199, tan-ettat
5285, vb. n. titacht, Wb. 25d 1 3, tetacht, Thes. i 496.26, tuttacht, 5/^2330, tauttacht
4420.

142
NOTES
[§§37, 38
The forms which occur in BB are 3 sg. pres, indie, clo-etet (§§37, 41, 43),
passive do-etegar (§42), pres. subj. do-eit (§45) and nom. pi. past participle
tetechtai (§§36, 39, 41, 54). Do-etet , do-etegar and tetechtai display an archaic
feature in their preservation of unstressed e after a non-palatal consonant — see
Introd. p. 13. In Class. O. Ir. the expected forms would be do-etat , do-etagar
(which is found in O’Davoren’s quotation of §42) and *tetachtai.
In the subjunctive one would expect do-et (or do-aut ) with non-palatal t,
rather than do-eit. Another compound of teit with an unexplained palatal t in
the subjunctive is in-otat (* in-oss-tet) with 3 sg. pres. subj. in-uait (the vocalism
is also irregular, Grammar §849) but 1 pi. in-otsam. The palatal t in con-eit , 3
sg. pres. subj. of con-etet (* com-in-tet) is regular and may have influenced other
/^/'/-compounds. One can compare also the alternation between -op Ml. 20b6
and -oip 42a2, both 3 sg. pres. subj. of as-boind ( Grammar §627).
In BB do-etet has the technical meaning of ‘ tracks (a swarm of bees) to the
place where it settles.’ This fits in with the verb’s more general meanings such as
‘ comes up to, attacks, catches, strikes.’ Another te77-compound which is used
with exactly the same technical meaning is in-etet (*in-in-tet): see note to §43.
A completely different interpretation of do-etet would be to postulate an
otherwise unattested do-etet (*to-in-tet). This would avoid the difficulty with the
pres. subj. do-eit, as do-eit would be quite regular cf. con-eit from con-etet.
However, as the meaning of *to-ad-uss-tet suits the BB examples perfectly well,
there seems to be no justification for postulating another compound.
i suidigetar. We emend -suidigther to -suidigetar, 3 pi. pres. ind. conjunct of
suidigidir (- ithir ) ‘ settles, alights ’, cf. §43 i suidigetar ( insuidegetar MS). In the
O. Ir. Glosses suidigidir is always transitive ‘ places, sets, fixes ’ (see D1L s.v.)
as in BB §§11, 14 rosuidiged, 50, 55 rosuidigthea. In BB it is also used intransitively
of a swarm of bees settling on a tree. The examples are §§37 i suidigetar ‘ where
they (the bees) settle ’, 43 frisa suidigetar ‘ on which they settle ’, 44, 45 i
suidigethar ‘ where it (the swarm) settles ’, 54 i suidigetar ‘ where they settle.’
In §42 the verb may be transitive or intransitive (see note).
This intransitive use of suidigidir seems to have been unfamiliar to most later
copyists and glossators. Thus the active endings sg. -ethar, pi. -etar have been
changed during the course of transmission to passive -ther (§§37, 44, 45). This
distorts the meaning of the text: bees settle on trees of their own accord and are
not placed there by any external agency. In the main A glosses originally
intransitive forms of this verb are generally taken to be transitive and passive,
and are explained by such glosses as 37f suidigthir no samaigtir iad ‘ they are
settled or established ’ cf. 44e, 45 s, 54b. However, gloss 43b i suidigend siad ‘ in
which they settle ’ on i suidigetar shows that the glossator here correctly took
the verb to be intransitive. Likewise in commentary in D (Appendix 3 (c)) the
intransitive tairisid (O. Ir. do-airissedar) ‘ settles ’ corresponds to suidigidir of
the text.
§38
do. For the construction dligim ni duit ‘ I am entitled to something from
you ’ cf. §4 dlegait-sidi deolaith doib, §6 ni dlegat in chethar doit-sin ni doib, etc. The
same construction is found in Welsh, see BBCS vii 364-6 ‘ dylyaf it: dligim duit .’
neoch ma. In ZCP xvi 270 Thurneysen lists ten examples of the use in legal
texts of neoch ma or i nneoch ma in place of the simple conjunction ma ‘ if’. With
one exception ( CIH 387.31 = AL i 176.2) all these examples are from BB, CU
and BFG (for the relationship between these three texts, see Introd. p. 28). The
other examples from BB are §§40 i nneoch ma, 42 neoch mani, 48 i nneoch mad.

§§38, 39]
NOTES
143
rna fo-ruasat orgguin. Our emendation of ma ruasasat to ma fo-ruasat (3
pi. perfective pres. subj. of fo-fich) is uncertain, and is dictated largely by the
sense. Though the phrase *fo-fich orgguin ‘ commits trespass ’ is unattested,
fo-fich is used elsewhere in the Laws with both fogal and cin : cia foghal friu
foruastar CIH 1114.11, ma forae neach cinuid nainceas CIH 1818.29. In §11 of
‘ a text on the forms of distraint ’ (Appendix 6 of this edition) cin, fogal and
or gain are all used to refer to grazing-trespass by bees. In B Comaithchesa ( CIH
69.14 etc.) the terms gaimorgain and gaimfuacht (■< fo-fich-) are used with the
same meaning of ‘ winter trespass.’
It is not easy to explain how an original ma fo-ruasat became ma ruasasat of
the MS. We suggest that the second -as- arose by dittography, cf. asasatairinider
for asa tairidnider in p. 26al0 of this MS (= Erin xvii 64 §1 ( CU )). For other
examples of dittography in BB see Introduction p. 2. For the omission of fo-,
one can compare the elision of the a of the preverb in cases such as masbo (= ma
as-bo ) CIH 493.32, madrodma (— ma ad-rodma) CIH 494.17.
Our emendation is not supported by the gloss madia roindsaigea siad, as
forms of the cpd. fo-fich are usually glossed by fuachtnaigid. Indsaigid indicates
saigid or ro-saig in the corresponding text (for examples see AL vi s.v. indsaigim).
One might therefore read ma rusasat orgguin a thire lit. ‘ if they reached trespass
of his land.’ Ro-saig is used of the extent of bees ’ ‘ grazing ’ in §8 noch is ed
ro-saig ind n-oircne (glossed ./. ind aired roindsaiges seig). However, although to
read ru-sasat would require less violent emendation, it would not make satisfactory
sense.
§39
The section on bees which settle on the branches of a tree, open ground or a
spread cloth belonging to a nemed includes § §39-41 . In §36 it was difficult to return
the swarm to the original owner because there would be objection to cutting the
trunk of the tree and it would not be easy to release the swarm in any other way.
Here there is no such problem since the bees have settled on a branch of the tree
or on open ground or on a spread cloth. The only difficulty is the capacity of a
nemed to protect those who come to him. By §42, however, the owner of the
land is no longer a nemed.
The heptad quoted here also appears in the collection of Heptads; see Introd.
p. 11. So far as it goes, J (H.3.17) agrees with A against H (Rawlinson B 487).
Unfortunately the evidence of J is lacking after fer airm d'eirg.
barr nemid. Barr is contrasted with §36 crann; for nemed see note on §15.
maigin. Cf. §42 bretha bairr 1 maigne. Maigen appears to mean ‘ piece of
open land ’ in this context; this meaning is not distinguished in DIL s.v., but cf.
mag 6 open land ’, ‘ large stretch of open land ’, as opposed to wood, mountain
or bog, and such phrases as maigen feirt CU §11. In origin maigen may be a
diminutive to mag (cf. the W. singulative -yn (m.), -en (f) <*-mo-, *-/>?«). If so,
‘ piece of open land ’ would be the original sense.
A swarm sometimes settles temporarily on open land i.e. grass or low-growing
vegetation (see Introd. p. 47) but more often on a tree, bush or fence, cf. glossd B.
brat scarthae. For the use of a spread cloth during the capture of a swarm,
see Introd. p. 47.

144
NOTES
[ §39
cain dimet ind nemid. Lucas’s glossf ( aisneidh leaf) shows that he took cain
to be 2 sg. ipv. of canaid ‘ sings, tells However, as this does not suit the context,
we take it to be the short form of the neg. interrog. cani ( Grammar §465). One
can compare B Crolige §25 ( Erin xii 20) cain comlai eslaine i mbroinn carna eich
fri fuile cauradl ‘ does not horseflesh stir up sickness in the stomach of wounded
heroes?’ (The glossator to this passage likewise takes cain to be 2 sg. ipv. of
canaid , as is clear from his gloss can-su ). Originally, both texts may have had
cani rather than cain here, as ni could easily have been mistaken for in.
An alternative explanation would be to read cain-dimet ‘ well do they (the
dignitaries) protect,’ as cain {cain) is often prefixed to verbs in O. Ir. (see DIL
s.v. 1 cain). This would make equally good sense in BB, but would not be possible
in B Crolige.
The ability of persons of high rank to give ditiu ‘ protection ’ is often
mentioned. One example is CIH 37.1 athgabail atagar i faiche neimed is coir dia
ditin; in another, CIH 361.6, discussed by Binchy, Celtica xi 25 n.26, one of the
trebaegal aigni is to bring cattle in distraint hi faithchi uasalnemid is tualaing a
diten. The classes excluded from that protection are illustrated by another
heptad {CIH 46-7 and 1847-8). A normalized text is as follows:
Ataat .uii. frithberta la Feniu i ndligiud ata annsam do airsciathud. Ni fil
cert do flaith na d’eclais na d’airechaib febe na snadud foisiten dia nditin:
(1) ditiu ar bithraith, (2) ditiu maic as-lui ara athair, (3) ditiu ingine as-lui
ara mathair, (4) ditiu mogo as-lui ara flaith, (5) ditiu manaig as-lui ar eclais,
(6) ditiu fir as-lui a recht De no duini, (7) ditiu mna as-lui a cain lanamno.
Ni con-araig Dia i tossuch na etarscarad duine.
‘ There are seven “ oppositions ” in Irish law which are most difficult to
shield. Neither lord nor church nor nobles have a right, nor is there
acknowledged safe-conduct, to protect them: (1) protection in the face of
a standing surety, (2) protection of a son who absconds from his father,
(3) protection of a daughter who absconds from her mother, (4) protection
of a slave who absconds from his lord, (5) protection of a manach who
absconds from his church, (6) protection of a man who evades the law of
God or man, (7) protection of a woman who evades the law of marriage.
What God has first joined together let not man put asunder.’ (Mark 10.9.)
Frithbert , literally ‘ opposition ’ is used of the offence by which a subordinate
or other person under obligation ‘ resists ’ his obligation and evades it or
absconds from a person who has rights over him {as-lui, elud). Nos. 1-4 and 7
involve obligations to persons, 5 an obligation to an institution or the head of
that institution, 6 more general obligations under law. Yet there is only partial
overlapping with the heptad of elodaig: no. 7 is identical with no. 5 in the list
of elodaig, nos. 2 and 3 partly coincide with nos. 6 and 7, assuming that the ben
and fer are counted separately. It is surprising that the slave who absconds from
his lord does not appear among the elodaig : in the Heptads he is the one most
often mentioned {CIH 17.18; 28.11-2; 29.3, 11). The elodach fine is the elodach
as-lui ara fine of CIH 38.31 and the ben as-lui a cain lanamno reappears CIH 39.6.
The unusual character of the heptad of elodaig in BB may stem from the
inclusion of the first two items: beich to-choislet i taidchu foindil. The second
may be a term for a human thief (cf. the glosses, the term cu glas for the alien
from outside Ireland and the term foindledach , e.g. CIH 38.30-1), but coming
as it does after beich to-choislet this is far from certain. As for the first item, it
seems probable that the original home of the heptad was BB: this alone would
explain the inclusion of bees and their position as the first item. If one supposed
that beich to-choislet was put as the first item only as a consequence of inserting
an already existing heptad into BB, one would face a serious objection: beich

§39]
NOTES
145
to-choislet is also the first item in the version in the Heptads. We must presume
that the compiler of the Heptads took it from BB, probably in the early eighth
century. If this is correct the clear divergence between the form of the heptad
in BB and that in the collection of Heptads would be explained.
Against this it could be argued that the inclusion of beich to-choislet among
the seven absconders is quite inappropriate at this stage of the tract. We shall
not reach bees which are elodaig until §44; in §39 it is still a question of beich
thetechtai, tracked bees; hence, perhaps, the use of the neutral to-coislea to mask
the distinction between tracked and un-tracked bees. Moreover, the bees in
this heptad could hardly be elodaig in the technical sense to be found from §44
onwards : according to BB a bee-keeper can never have asserted any claim to a
swarm of un-tracked elodaig. A nemed would therefore not need to provide
ditiu. Though these bees are among the seven elodaig they must have been
tracked, and therefore precisely not be elodaig in the sense proper to BB. One
might, therefore, claim that the Heptad is a later insertion into BB.
Such a conclusion would be premature. Because as-lui is used in a special
sense from §44 onwards, we are not thereby compelled to conclude that it could
not have been used in a wider and more normal sense earlier in the tract. There
is no cogent reason for dismissing the heptad as an insertion. True, heptads as
a class may be relatively late compared with triads, but they may well have been
coming into use by lawyers in the second half of the seventh century, early enough
for the date of BB. Finally, there are two heptads in the sister-tract CU which
suggests that this form became popular at an early date in the school from which
these tracts derive.
We write -dimet with a short i though Thurneysen, Grammar §831, writes
-dimea with a long /. The exx. in the Glosses do not mark the /' long; moreover,
three exx. of di-ella in the Glosses have a short i in the prototonic (see DIL s.v.
do-ella).
doda-airret. We emend dodauret A to doda-airret; the substitution of aurfor air- is common in this MS — see Introd. p. 3. The cpds *to-air-reth- and
* to-ad-reth- seem to have become confused in Mid. Ir. MSS (see DIL s.v. do-airret
and Grammar §822). Consequently, the original of BB may have had doda-airret
(from *to-air-reth-) or doda-aret (from * to-ad-reth-).
eludaig. A has eludaig here but eolodaig later in the same paragraph. H’s
spelling elaidthi provides yet another way of spelling the second syllable. The
evidence suggests that the vowel of the second syllable was sometimes long,
sometimes short (see DIL s.v. eludach). Moreover, even when the vowel was
long, there was still variation between a 6 and u in words of this type (adjectives
in -ach or -ech derived from verbal nouns of A III verbs). A good example is
impaidach Sg. 60a8 from impud/impod , though it is difficult to see how a long
a arose. Since the vowel of the second syllable of these words seems to vary in
quantity and quality, we have not supplied a mark of length, and have retained
A’s variation in the quality of the vowel.
na dim nemed De na duini. Cf. §§28 na deni, 39 (last line) na dama. J here
reads nadh dimh, and at the end of the paragraph K has nad dama, but there is
no need to restore the final dental of nad against the testimony of most of the
MSS, cf. Ml. 23c20: . . .fri nech nadeni ole friut ‘ to someone who does not do
evil to you ’. Here assimilation of final and initial lenited d has yielded d,
presumably [d] in accordance with Grammar §231.3. The H variant -dime may
possibly go back to an original pres. subj. -dimea, cf. §40.

146
NOTES
[§§39, 40
The H. version of this heptad has nach nanrachtaig after duini, but these words
are not present in A or J. In §40 H has nac nanrechtaig after duini, again disagreeing with A (J breaks off before this point). The same phrase is found in
Eg. 88 f. 46a (= CIH 1402.39-41) : ar ni dim neimed de na duine nach nanrechtaigh
forai ‘ for the dignitary of neither God nor man protects any criminal ( anrechtaid )
who has committed an offence (3 sg. perfective pres. subj. of fo-fichT). Here nach
nanrechtaigh forai is glossed inti his for amhdiartaid gaite *j braite 6 he who is
engaged in illegality of theft and plunder ’.
beich to-choislet. We restore docoislet of the MSS to to-choislet. This is
suggested by the ‘ etymological * gloss toich aslaiet by Cairbre, the third A
glossator (see Introd. p. 14).
taidchu foindil. See note on cain dimet ind nemid above.
fer airm deirg. This appears to be the man who has just committed homicide
and whose responsibility is undeniable. Cf. deirgfine , CIH 429.26, for the kinslayer, and the old gloss CIH 917.28; fer laime derge ( DIL , L, col. 37.11-20),
Welsh Uofrudd (see D. Jenkins (ed.) Llyfr Colan, n. on §144). The glosses in J
and H may perhaps be correct in claiming that a man remains a fer airm deirg
until the wergeld has been paid; but, if so, it would be strange that no nemed
could afford him protection, even if he were willing to pay compensation and
his kindred could help him to pay. An alternative explanation is that the fer
airm cleirg is an habitual killer, who may even have been ‘ proclaimed ’ by his
kindred. This would account for the inability of the nemed to give ditiu.
ben as-lui a cain lanamno. The third glossator in A has used a phrase from
another gloss, CIH 233.24, on ben deiraig a lanamnus, possibly without understanding it fully.
mad nech na dama coir iar setaib taide. Cf. the early gloss in K which suggests
that this person has gone into exile secretly evading his obligations. This,
however, is too literal an interpretation of iar setaib taide : cf. B Crolige §63 and
note ad ioc.; CIH 40.21-2, tech tadut nacon daim cert na dlige[d] do duine. The
condition is stated generally CIH 534.27-8: . . . muna clena a mac a gaire acht
mad athair anfoltach. On the duty of goire see D. A. Binchy, Celtica iii 228-31.
§40
In this sentence the differences between the two textual traditions, that of
the Heptads and that of BB, are at their most troublesome. For dime H see note
on §39 na dim, and for the addition nac nanrechtaig in H see end of same note.
ara-fogna. A and B (namely the textual tradition of BB) have oirngne and
oirgne against arafogna H. For the spelling oirngne cf. §8 where A has the same
spelling for the gen. sg. of or cun, and such forms as do-ringensat, Ml. 16d6,
explained by Thurneysen, Grammar §681, as due to the influence of a following
nasal. The textual problem was noticed by Atkinson, AL vi 723. He interpreted
oirgne as acc. pi. meaning ‘ spoils ’, but this sense of the word is not attested
before the late Mid. Ir. period. There is also an early gloss in B. His explanation
is that the thievish parents may have young children who, in the absence of an
adult brother or sister, would be likely to die of want and hunger. If this gloss

§§40, 41]
NOTES
147
applies to the text as preserved in B, duine oirgne must have been understood by
the glossator to mean a person liable to die for lack of care. This, however
sensible an explanation in itself, seems rather unlikely as an interpretation of the
phrase didne oirgne. Another problem is the relationship of this phrase to the
cpd. duinorcun ‘ homicide If one were to emend to duinorcnea this might
accord better with the explanation in the B gloss. It is perhaps possible that an
old set phrase consisting of a preposed gen. + noun might survive as late as the
seventh century, even in straightforward prose, but in fact the comparison,
however problematic in detail, with Welsh dyniorn suggests that the old term was
the cpd. duinorcun and not any phrase such as *duini or cun.
In effect, if we follow A and B it is possible to think of explanations but none
of them is sufficiently likely to justify adopting their reading as against H’s
arafogna. The maxim of the lectio difficilior should, however, prevent the reading
oirgne from final rejection.
The use of the cpd. ar-fogni in the context of go ire is natural and can be
paralleled: Irishes Recht i. Dire §26, each manach arda-fogna mamaib gaire
‘ every manach . . . who ministers to it (his church) according to the obligations
of piety.’ Ar-fogni is distinguished in meaning from fo-gni in that it usually
means ‘ ministers to ’ rather than simply ‘ serves it may, in particular, be used
with biad as object and mean ‘ prepares (food)’. Hence it is entirely appropriate
when used of a person fulfilling his duties in respect of goire. The subject of
ara-fogna must then be the person bound by the duty of goire. If his parents are
prone to denying the rights of others and to theft, he is exempt from goire , but
not if he should leave behind him a duine to whom he should minister. At this
point we may return to the gloss in B, according to which the duine would be a
child, or children, who, if goire is not performed, ‘ may suffer a death of want
and hunger.’ It is an important argument in favour of arafogna that the early
gloss in B still makes perfectly good sense — indeed, it provides the only plausible
explanation. The gloss in B — probably 9th century — -is, therefore, no proof that
the text on which the glossator was working had oirgne rather than arafogna.
A quite different explanation would be to take arafogna etc. as a prepositional
relative clause, and to translate ‘ instead of whom he should serve.’ His father
might be a client or other dependant and owe his lord fognam which he was not
able to perform. Initially this interpretation seems plausible, but it is not likely
that a thieving, non-law-abiding parent would still be in such a relationship.
It is much more likely that we have here the cpd. ar-fogni. Cf. Thurneysen,
ZCP xvi 271 11. 19-24.
frisimbi coir do tintuth. The sense of this clause is clear but syntactically it
may be taken in two ways. Cf. CG 505-6: cis lir ata chorai do rig do giull fora
thuatha where two phrases with do follow coir. The first defines the person whose
actual or possible conduct is described as coir , the second specifies the conduct
which is coir. We may therefore write frisimbi coir do tintuth, in which case
coir do is paralleled by chorai do rig ; or we may write frisimbi coir do thintuth
where coir do thintuth corresponds to chorai . . . do giull. Cf. also CG 553, Cia
as choir *] as techtae do clenum biid rig. The first construction is also well instanced
in the O. Ir. Glosses (see D1L, C, col. 314.19-36).
41
glossb. In B co cend mbliadna is glossed .i. rannait i suidiu etarru, no marbaid
madh ferr leo, no bit etarru dana beos ‘ they divide between them then, or they

148
NOTES
[§§41,42
kill if they prefer, or they (the bees) still remain between them.’ At 44d co cenn
teora mbl- is similarly glossed by B ./. marbaid no rannait ~j (? read no) bit etarni
beoss.
It is difficult to reconcile these three alternatives with the procedure laid
down in the text of §§41 and 44. Rannait in both glosses no doubt refers to the
division of honey between the original owner of the bees and the land-owner.
It can hardly refer to a division of the bees: a colony normally has only one
fertilized queen, so a share of the bees would be of no use unless there was a
queen among them.
Marbait could be taken to mean that the bees are to be killed at the end of
the first year (as in §41) or at the end of the third year (as in §44). However,
this would conflict with the evidence of the text and its commentaries. In §41
the bees become the sole property of the original owner after the first year (is 6
suidiu a n-imchomet cf. Appendix 1 (c) is diles dfir na mbeach otha sin amach iad).
Similarly in §44 the bees become the sole property of the land-owner after the
third year (is in tire i suidigethar is coir bunad a tuisten cf. App. 1 (c) is diles dfir
in craind iad ota sin amach). It is implicit in §§19-22 (see Introd. p. 43) that the
bees were normally not killed at the extraction of the honey from the hive. See,
however, note to §27 collud.
The third alternative bit etarru beoss ‘ they still remain between them ’ (i.e.
they both retain a share in the ownership of the bees) seems also to contradict
the text, as in §41 the bees become the sole property of the original owner after
a year, and in §44 they become the sole property of the land-owner after three
years. In CIH 192.32 = AL iv 102.17 (in the tract on comingaire ‘ joint grazing ’)
bith mart an aighe atarrtu means ‘ the flesh of the bullock is between them ’ i.e.
is divided between them. However, in our B gloss bit etarru cannot mean ‘(the
bees) are divided between them ’ for the reason given under rannait above.
To conclude, it would seem that the B glossator took his glosses here and at
44d from some other context, and applied them inappropriately to bees.
is deolaid a trian ticc a nnemed. The parallel passage in §37 has is deolaid a
trian tic fer mbunaid where the object of the verb is a person. This raises the
possibility that we should emend annemed of the MS to in nemed, as nemed
seems normally to be a masc. o-stem when used of a person (see note on §15
nemed).
However, BB does not always express a clear distinction between a holding
of land and its owner. Thus the literal translation of §37 a trian n-aill do nemud i
suidigetar, a trian n-aill do thir do-melat is ‘ the second third [of the honey]
comes to the nemed(- land) in which they settle, the other third comes to the land
where they feed.’ The meaning is clear: ‘ the second third comes to the nemedperson on whose land they settle, the other third comes to the owner of the land
where they feed.’ The neuter noun tir ‘ land ’ thus stands for ‘ the owner of the
land.’ Similarly the neuter a nnemed ‘ the nemed- land ’ may stand for the nemedperson who owns the land.
Another argument against taking annemed to be a scribal emendation of in
nemed is the fact that the O. Ir. def. art. in is not modernised to an elsewhere in
A’s text of BB.
§42
In both A and B the initial letter is a large capital, as if marking the beginning
of a fresh tract. However, as §42 is closely linked with the preceding paragraphs,
this is certainly a scribal error: see Introd. p. 31 (note).

§42]
NOTES
149
glossb. In B ./. bitt suidiu do ni gaibther glosses 7 maigne. We take bitt to be
3 sg. consuetudinal present of atta, suidiu to be for i suidiu ‘ seated, settled ’ (or
‘ there ’), and do to be for dano ( dono ) ‘ indeed.’ There may originally have been
two separate glosses here : ‘ they are settled indeed ’ (referring to the bees) and
‘ that which is settled upon ’ (referring to the maigen ‘ open land ’).
glossc. We have not been able to translate the gloss .i. di neoch ateagear
daigh dibh go aith by the first glossator. The form ateagear clearly corresponds
to do-etegar (3 sg. pres. ind. pass, of do-etet ) in the text, but we do not know how
to interpret it. It may be from an artificial compound of ad- and teit or a mistake
for eteagar (the original composer of the gloss having simply glossed do-etegar
by omitting the first preverb : ete(a)gar). It is surprising that do-etegar was not
glossed by some phrase containing coimitecht, vb. n. of con-eitet ‘ accompanies
as the three occurrences of do-etet in the text are glossed bis ina coimidecht by
the first glossator.
The rest of this gloss daigh dibh go aith is obscure and probably corrupt.
Go aith ‘ quickly ’ usually glosses a{i)th- (occasionally ad-) but there is no trace
of either preverb in the text.
The fourth glossator adds two further glosses: ./. do neoch dib dochum a
ninaitcither a tiachtain in the right margin and .i. doinchuitchet in the left margin.
Elsewhere in BB inaitchid [*in-ad-tet] is used to gloss in-otat ‘ enters ’. Thus in
44a and 45d the first A glossator uses the 3 sg. perf. ra inaitcestar to gloss in-otat.
Here, however, the context seems to demand some such meaning as ‘ tracks,
follows ’. Possibly the fourth glossator is thinking of O. Ir. in-etet , used of
tracking bees in §43 (glossed ineitces by the first glossator).
The fourth glossator’s other gloss .i. doinchuitchet seems to be 3 pi. pres. ind.
of * to-in-com-tet ‘enters together (?)’. This cpd. is otherwise unattested and
perhaps not genuine.
B’s gloss o bechaib ‘ of bees ’ refers back to di neoch of the text, cf. la nech
huain Wb. 6b20. It cannot mean ‘ by bees ’ (agent).
arddnemed. In BB there appears to be no distinction in meaning between
the terms nemed, uasalnemed and arddnemed. See note to §15 nemed (3).
ailid. Following Thurneysen ( ZCP xvi 272) we emend ailed to dilid, taking
the -id to have been altered to -ed through contamination with the preceding
nemed.
son. Here we take the neuter anaphoric pronoun son to mean ‘ that case,
that situation ’ and to refer back to the entire heading Bretha bairr 7 maigne di
neoch do-etegar, acht arddnemed. Cf. note to §2 dlid-side.
mani-ro foiscet. The MSS have mani rofoiscet (with / erased) A, mana
rofoisccet B. We read ro foiscet which we take to be 3 pi. perfective pres. subj.
of fo-scoichi [*fo-scoch~] ‘ moves, shifts ’. This identification is supported by
the B gloss .i. foscugud (vb. n. of fo-scoichi ).
We take the form foiscet to have arisen through reduction of the verbal stem
scoch- to sc- after syncope. One can compare deuterotonic di-rosci Ml. 57al
‘ surpasses ’ «* de-rosc’’ chi <C *de-ro-oss-scochi) but prototonic -derscaigi Sg.
40a 17 ( Grammar §137 (3)).
Elsewhere this reduction has not taken place in O. Ir. Thus conosc(a)igi
* moves ’ [ *com-uss-scoch -] has 3 pi. pres. ind. conosciget Sg. 65b9, 3 sg. fut.
connoscaigfe Ml. 61dl. In legal MSS, however, the reduction is well attested,

i5o
NOTES
[§42
e.g. 3 pi. pres. ind. conoiscet CIH 1483.10; conoiscead (gl. ./. cumscaidit) CIH
239.13; 3 sg. pres. subj. conidnoisce CIH 581.26; 3 sg. fut. conidnoiscfe CIH
2316.2 (j Eriu xx 46 §37). It seems unlikely that forms in -scig- are to be restored
in all these cases. We therefore do not emend to mani-ro fosciget.
We are unsure of the correct interpretation of neoch mani-ro foiscet crick
‘ unless they (the bees) have shifted territory ’. This paragraph deals with tracked
bees which have settled on the branches of a tree or on open ground in the
property of a no n-nemed land-owner, who gets one quarter of the honey for a
year from the bee-keeper. As has been described in the Introduction p. 47, when
a swarm of bees emerges from the hive it usually settles for a few hours on a
nearby tree or bush, while scout bees are choosing a permanent nesting-site.
The compiler may have added the stipulation neoch mani-ro foiscet crick to
meet a situation of the following type. A swarm belonging to X might have
settled temporarily on property belonging to his neighbour Y. Although X had
followed his swarm into Y’s property, it moved off again before he could capture
it, and went to a permanent nesting-site in a tree on Z’s property. In this case
Y would have no claim to a share in the honey. A similar situation is dealt with
in Recension E of the Latin texts of the Welsh Laws (ed. Emanuel p. 483): si
heit alicuius in arborem alterius hominis intraverit, et possessor arboris earn
recipere non permiserit, licitum est possessori apum post annum et iinum diem
suum heit capere, ita tamen quod met in duas partes inter eos dividatur. Examina
tamen si qua ex arbore exierunt, possessor arboris non habebit (translated in
Appendix 7, p. 199). The fourth A glossator may have the same idea in glossf
when he refers to the bees being isin crand asin crand which seems to mean ‘ in
and out of the tree.’
If our interpretation of neoch mani-ro foiscet is correct, it is strange that the
compiler does not add it also in §41.
autsad. The original meaning of autsad seems to be ‘ store-house, treasury ’
(glossing thesaurus Ml. 51d8) but it is also used in a wider sense: ‘ abode, lodge,
apartment ’ ( DIL . s.v. etsad ). Its etymology is obscure. Thurneysen ( Grammar
§80 (c)) takes its first element to be ad-uss-, but makes no suggestion about the
rest of the word. The late Prof. David Greene told us that he had retracted the
suggestion which he made in Fingal Ronain and other stories (note to 1. 558) that
autsad is from *ad-sossad, but he was still inclined to see sossad ‘ abode, etc.’
as the second element (fad-uss-sossadl ). He compared domsod ‘ dwelling-place ’
from dom ‘ house ’ + sossad.
The precise meaning of autsad in BB is unclear. The general sense of dligid
each tir autsad di neoch suidigther fris seems to be that every land-owner is
entitled to a share in anything valuable which may settle or be deposited on his
land. Apart from bee-swarms, this principle may have applied to valuables
deposited by wind or flood. It may also have applied to articles buried or hidden
on another’s land, cf. ALIW i 552 §30 which states that anything hidden on
another’s land belongs to the land-owner, except for a hoard of gold (which
must be handed over to the king).
Apart from BB, the only legal attestation of autsad which we have found is
O’Dav. 101 (CIH 1469.30) cidh forsa sui[di]gedh techta cana tafuinn do autsudhaibh
tuaithel ‘ on what is the due of the law of hunting fixed for treasure-stores (?) of
the tribe?’ This may refer to the right of a land-owner to a share of any wild
animal killed on his land. In Welsh law one can compare ALIW i 736 §1 which
deals with the division of a stag between the huntsman and the owner of the land
in which it was killed.

§§42-44]
NOTES
151
On the basis of the B gloss to our text ./. log autsada , Dr. Binchy suggests
that autsad may possibly be a technical term for the ‘ storage fee ’ due to land
or buildings for ‘ enclosing ’ i.e. sheltering objects.
suidigther. We read suidigther (following suidigthir MS) ‘ is deposited ’
rather than suidigethar ‘ settles ’. When used of bees, suidigithir is always intransitive in BB (see note to §37 i suidigetctr ). Here, however, the rule does not
apply to bees only (see note on autsad above), so we do not feel justified in
emending to suidigethar.
§43
in-etet. This cpd. ( <*in-in-tet ) seems to have exactly the same meaning as
do-etet (see notes to §37). Emendation to do-etet is out of the question, however,
as the preverb is confirmed by the gloss ineitces. The only other attestation of
in-etet is CIH 427.3 (— AL v 516.5) fer inetet toin a mna tar crich ‘ a man who
follows the backside of his wife over a boundary.’ Here Thurneysen (Jrisches
Recht p. 64 §4) suggests emending to in-otet (= in-otat [*in-oss-tet]). However,
its meaning ‘ enters ’ suits neither context. In BB the cpd. in-otat occurs in §§44
and 45.
suidigetar. We suggest that both cases of suidigidir in this paragraph are
intransitive. If not, suidegetar and suidigethar of the MS must be emended to
suidigter. See note to §37.
gloss 43c. The reference here to ‘ the fourth hostile territory not beyond an
arm of the sea ’ is mere pseudo-learning on the part of the glossator, and has
no relevance to the text. This phrase is quite common in glosses and commentary,
e.g. CIH 295.33 (= AL iii 318), CIH 228.20 (= AL v 442).
do fiur. The main scribe omitted these words by jumping on from the do of
do fiur to the do- of doda-etet. The third glossator inserted dofir above the line;
his correction is confirmed both by the sense of the passage and by the gloss don
fir of the first glossator. For a similar correction by the third glossator, see §8
ro-saig and note.
oa n-elat. Our emendation of oatelat to oa n-elat (3 pi. pres. ind. of as-/ui,
-e/ai [*ess-lu~] ‘ escapes ’) is supported by the gloss o nelan siad and also by the
reading oanelat in §44. We take the intrusive -t- to be a scribal slip, perhaps
suggested by the cpd. do-ernai, -ternai ( *to-ess-ro-sni -) which likewise means
* escapes.’ There may also be influence from the rare cpd. do-elai , -telai (*toess-lu-) ‘ escapes.’
§44
This and the following paragraph treat of the situation in which the claim of
the original owner of the bees is at its weakest. Not only have the bees swarmed
without being tracked by him, but they have not been tracked by anyone else
(as in §43). Nevertheless the original owner can swear that the swarm entered
his neighbour’s land at the time of the day when swarms might be expected; he

152
NOTES
[§§44, 45
must be assumed to have seen them leave his hive but have been unable to keep
up with them. He thus at least knows the direction in which they have swarmed.
The fourth glossator (under glossa) has a different explanation which relies on
the phrase i n-inbuithib cuir saithe : for him the original owner can testify that
the bees are his, but he cannot testify which of his bees they are. He cannot
correlate the swarm on the land of his neighbour to one particular swarm as it
left the hive apparently because more than one swarm was leaving his apiary at
much the same time. But this is the situation of §45, not §44. The glossator’s
arithmetic will become clear if one compares this gloss with his gloss 42d. There
the division was f and i; here it is ^ and \\ the difference, as he says, is due to
the doubt raised because the bees have not been tracked.
ro-ch-lamethar forgall. The use of the punctum delens under -ch- was
presumably subsequent to the writing of the standard gloss on -ch-, seichim no
indsaigim . . . ; but it does illustrate the usefulness of this gloss since, in the next
stage of copying, -ch- would probably have been omitted (for this and other exx.
see Binchy, Celtica v 83). For MS forgull see third note to §34. For the MS’s
lainethar for -lamethar, cf. motat for inotat in the same sentence.
a cheli. Cele is, in this context, synonymous with comaithech : cf. §45 and the
tract on comingaire, CIH 192.1-33; 576.25-577.24. In the second, and better,
MS ceile is used for a neighbour co-operating in joint-herding at 576.28; 577.20,
22, but comaithech at 577.7.
i n-inbuithib cuir saithe. Cf. Introd. p. 46. Cora of the MS is no doubt the
later u- stem gen. sg., as in Irish Grammatical Tracts, Declension §70. One can
compare Appendix 1 (b) aimsear cora saithe.
gloss 44c. The lenited initial of fodailet suggests that the glossator has omitted
is cain. The normal ‘ etymological ’ gloss on con-fodlat is is cain fodailet, cf.
17b, 45e.
each torad is in opposition to saithe and makes it clear that it is not the swarm
itself which is to be divided (impossible since there is only one queen), but the
produce. B’s gloss, under d, seems wide of the mark in referring to the possibility
of killing the bees: see note to §41t).
is coir bunad a tuisten. B has preserved the older construction with is rather
than rel. as: see Grammar §506.
We have been unable to understand B’s gloss is aim is coir domsud na mbech
do buith occ lus. When used in a local sense, the prep, oc normally means ‘ next
to, beside, at ’ (see DIL s.v.). However, it is difficult to see why the B glossator
would claim that the proper abode for bees is beside a bush (hardly for shelter?).
It is true that swarms frequently settle temporarily on a bush, before the scout
bees have found a suitable nesting-site (see Introd. p. 47). The B glossator refers
to this in 39d argaib-som sonnach no lus ‘ it (the swarm) settles on a fence or
bush.’ But his use of the word domsud here in 44e implies a permanent abode,
as in gloss 18a.
§45
The syntax of this paragraph is of unusual complexity (cf. Introduction,
p. 36). The entire paragraph is one conditional sentence of which the protasis,
with subordinate clauses, runs from Mad to comoccus doib uilib and the apodosis

NOTES
153
[ §45
from con-fodlat to int sainreduch. The legal situation is however, clear: it is
undisputed that the swarm in question has come from the place where there are
many hives: what cannot be shown is from which hive it came. Not merely has
it not been tracked, there is no one who can swear an oath so as to claim it.
nocli as-rulai. We emend arrulla (gl. gur ela) to as-rulai, 3 sg. perf. of as-lui
( *ess-lu -). The MS shows a misreading of -rr- for -sr-, and the common doubling
of the / (see DIL s.v. as-lui ), presumably influenced by verbs with initial si-.
We take the ending to have been -ai in the original text, cf. asrochumlai Sg. 7bl9,
3 sg. perf. of as-comlai ( *ess-com-lu- ); conhiialai, Thes. ii 320.7, 3 sg. pret. of
con-uala ( *com-uss-/u- ). On the other hand, the ending -ae occurs in asruchumlae
MI. 17b2. See Grammar §679 and Greene, Erin xxvii 32.
Another course of action is suggested by Dr. Binchy ( Celtica v 90). He
points out that one might expect a prototonic form of the verb after noch, cf.
Erin xx 44 §35 = CIH 2315.20 noch ( nocha MS) dlig ‘ and he is entitled to
Consequently, he would emend here to noch-erla{i).
nach laimethar. This is presumably the negative to ro-ch-lamethar (§44),
as suggested by Thurneysen ZCP xiii 299. If so, nach appears to derive from
*na-kwe since *na-k (W. nag ) yields na. The exx. collected by Thurneysen,
Grammar §866, are, however of naich or nach in what appear to be copular
sentences: e.g. naich in naich diless du[a]is diupartee, Biirgschaft §81, may be
from *nak-ess rather than directly from *ne kwe as suggested by Watkins, Celtica
vi 10. Yet, just as ni «*mss <.*ne est ) is used before verbs, replacing *ne, so
may the use of nach ( <C*nak-ess ) have been extended. Elsewhere na and nach are
in complementary distribution, Grammar §§862, 863 and 865. The difficulty is
that one might have expected this complementary distribution to extend to this
present construction, yielding na laimethar. But since the different forms of the
negative other than ni have all been appropriated to some particular construction
or constructions narrower than their original function it may well be that nach
was used to co-ordinate negative clauses. Certainly nakwe was not entirely
ousted by *nak since it is reflected in such constructions as nacham-dermainte
Ml. 32d5 ( Grammar §419) where nacham- derives from *nakwe me: see Watkins,
Celtica vi 10. As Watkins argues the distribution of -ch in O. Ir. — only after
proclitics — must be a relatively recent innovation, since, for example, the proclitic
status of ha in ba-ch ri Temro is itself recent; similarly the distribution of nad as
opposed to nach is probably late in the pre-history of Irish {ibid. p. 26). One may,
therefore, explain nach in nach laimethar as either (1) from *nak-ess , ultimately
*ne kwe est, or (2) from *nakwe <C *ne kwe. The first explanation appeals to the
parallel with ni, the second to the survival of *nakwe before infixed pronouns
beginning with a consonant, and to the relative lateness of the distribution of
nach, nad and na. There is no need to emend our text to na laimethar.
B’s gloss 0 reasserts the old doctrine about the fiadu (see Binchy, TPS, 1959,
23 = CLP, p. 119), quoting, in the process, an archaic maxim.
in-otat saithe do-eit. For do-eit see note to §37 do-etet. The syntax at this
point is most complicated. There seem to be two ways to construe in-otat saithe.
First, it may be taken as co-ordinate either to noch as-rulai and nach laimethar
or to bed lais even though it lacks a connective. Secondly, it may be taken as a
nominal clause and as the subject of bed. With the first explanation the meaning
will be straightforward: it is simply a further condition that the swarm enters
the land of a man who is neighbour to all the owners of the hives in that place,
comoccus doib uilib. The second explanation is more difficult: here the oath

154
NOTES
[§§45, 46
which is not sworn would be to the effect that the swarm, which the swearer was
in a position to track, do-eit, entered the land of the neighbour with him, lais.
Here lais would refer not to ownership but to the physical proximity of a tracker.
A comparison with §44 argues in favour of the first explanation and in favour of
taking in-otat saithe as co-ordinate with noch as-rulai and nach laimethar : the
owners of the bees know (a) that the swarm has entered the neighbour’s land and
(b) that it comes from one of their hives. The only point for which none of them
can swear is to which of them the swarm belongs. The absence of a connective
argues for the second explanation. With some hesitation we have opted for the
first.
hi tir. The preposition i is here used exactly as do is used elsewhere in the
text (e.g. §§37, 41, 43).
forgall na airthech. For the MS spelling with -u-, see note to §34 nadidlamethar airthech na fortach. The third glossator suggests that the forgall is
the oath of the neighbours, whereas the airthech is the oath of the land-owner
(‘ the owner of the branches ’). However, these are just guesses on his part.
For forgall, see CG, Legal Glossary, s.v. aire for gill ; for airthech see note to
§34 above.
§46
fo-gaib. We read fogaib with D against fongaib A. A’s infixed -n- cannot be
correct, as in O. Ir. the nasalizing relative clause can only be used in place of the
leniting relative clause when the antecedent is felt as the object of the verb
0 Grammar §501).
A also has an intrusive -n- when this phrase recurs at the beginning of §47,
but in §48 it has fogaib without -n-. In §49 it has fodgaib (with contamination
from §§46, 47, 48 fod{ti)gaib with infixed pronoun) but the correct reading fogaib
is present in L.
frith mbech, lit. ‘ a finding of bees In O. Ir. both frith and frithe are used
for ‘ something which is found, a waif, an estray ’ and there seems to be no
distinction in meaning between them. Frithe [m./o] is a past participle formation
from fo-gaib (cf. Grammar §§691(d), 763) and Dr. Binchy suggests in the
vocabulary to CG that frith may be an old verbal noun. The gender and
declension of frith is uncertain and we provisionally take it to be a neuter o-stem,
cf. note below to fod-gaib. Outside BB, the only legal occurrence of frith is CG
257 frith faithche where the context is difficult (see note). Heptad lxiv {CIH
55.18 = AL v 320) lists the seven frithi of Irish law, which include frithe fai[th\che
‘ the estray of a green ’ as well as frithe ro(f)ida ‘ the estray of a great wood ’
(cf. BB §49 frith mbech hi ruud ‘ an estray of bees in a great wood ’).
faithche. This word is usually translated ‘ green ’, which is slightly misleading,
as it suggests a large uninterrupted area of grass. This was clearly its meaning
in cases such as for sin faichthi ina Temrach , RC xxiv 190 ‘ on the green of Tara ’
(for further examples see DIL s.v.). In our text, however, faithche is used to
describe the area around a typical farm-house, presumably including the best
and most carefully managed part of the farm, i.e. the ‘ in-field It would include
land used for pasture, as in CG 198 where the bo-air e is said to have a faithche in
which there is always sheep. It would also include land used for tillage as in
B Comaithchesa ( CIH 72.18 = AL iv 108.10) where a fine is laid down for
trespass by a pig in a gort faiche, probably an inclosed corn-field.

§46]
NOTES
155
The faithche is felt to be owned more exclusively than the area which is
described as sechtar faithchi (§48). Consequently, the finder of a stray swarm
gets one third of its produce if it is sechtar faithchi, but only one quarter if it is
in a faithche. BET s definition of the extent of the faithche as being ‘ as far as the
sound of a bell or the crowing of a cock reaches ’ is not found elsewhere in the
text of the Laws (though repeated in commentary, e.g. CIH 57.18; 1753.11).
Also in commentary (CIH 51.15 = AL v 328.8) the faithche is defined as na ceithri
guirt is nesa don baili ‘ the four fields which are nearest to the house ’ but the
evidence of BB shows that in Old Irish law it embraced a larger area than this.
In general, short measurements of length, breadth, height, etc. are quite
precise in the law-texts (as in the description of proper fences, ditches and walls
in CIH 73.7-18 = AL iv 112). Longer measurements are vaguer, and the use
of numerals is avoided. Thus in BB §8 the extent of a bees’s grazing is defined
‘ as far as a cow reaches on pasture until milking-time ’ and in CG 219 the airlise
is said to extend the length of a [spear] cast all round the house (for other
examples of this use of airchor ‘ a cast ’ see DIL). In Welsh law (ALIW i 94) the
wife of a taeog is allowed to lend only her sieve, ‘ and that as far as her voice
can be heard from the dunghill requesting its return ’. One can compare also
the modern use of 6 a stone’s throw ’, ‘ an ass’s roar ’, etc. as approximate
measurements of distance.
guth cluicc. The reference here to ‘ the sound of the bell ’ suggests a monastic
environment. The 8th century Felire Oengusso (Feb. 13) preserves a tradition
that honeybees were introduced to Ireland by a monk named Mo Domnoc.
However, the linguistic evidence suggests that bee-keeping is much older than
Christianity in Ireland (see Introd. pp. 40-42).
fod-gaib. The gender of frith is uncertain. The MS has fodgaib here (indicating
that it is neuter) but fodngaib in §48 (indicating masculine). We provisionally
take it to be neuter, reading fod-gaib in both cases.
hi fogbaither. This verbal form must be examined together with §47 hi
fogabar and §48 hi fogbaither (fogbaiter MS). In all three cases the context
requires the 3 sg. pres, indie, passive (prototonic) of fo-gaib. Here hi fogbaither
is glossed i faghabar he, in §47 hi fogabar is glossed / fagabar e and in §48 i
fogbaiter is glossed / fagar e (confirming that the text originally had a singular
verb).
The main problem is whether the variation between forms with and without
-th- could have been present in the original text. Bergin (Journal of Celtic
Studies i 184) points out that the passives -dleg(h)ar and -dligther occur in consecutive paragraphs of Berrad Airechta, an early tract preserved in H.3.18
(Thurneysen, Biirgschaft p. 17 §§52, 53). In such a case, however, the variation
in spelling could be scribal, reflecting the apparently indiscriminate use of forms
with and without -th- in Mid. Ir. e.g. co[n]ggarar Trip. 462, congairther LU
5348.
From the few examples of the pres, indie, passives of ^tf/6-compounds in the
O. Ir. Glosses, it appears that when the verbal stem is stressed, one gets -th- as
in the simplex (con-gaibther Wb. 21c6, do-gaibther Sg. 28b20) but when it is
unstressed, there is no -th- (-tuargabar Wb. 14b22, do-furcabar Sg. 43a3, conocabar Mb. 57d8). Thus, on the evidence of the Glosses, hi fogabar is likely to
have been the original reading in §§46, 47 and 48 of our text. However, as the
Glosses may show a later more uniform state of development than BB, we have
thought it best to keep the MS readings.

156
NOTES
[§§47, 48
§47
This paragraph deals with an exception to §46. A man finds a colony of bees
in a tree in a faithche belonging to someone else. If the colony has been in
existence undetected for more than a year the landowner gets half of the produce,
rather than the regular three-quarters. As the third glossator points out (glossc)
this is to penalize him for ‘ not making a circuit ’ of his property. If this rule
was ever applied in practice, it must have been difficult to establish whether a
colony had been present for more than a year or not. In favourable conditions
bee numbers can increase rapidly, so a large colony with considerable stocks of
honey might be only a few months old. One possible indication of the age of a
colony is the condition of the honey. If much of the honey in the combs was
granulated, this would prove that the colony had existed for a considerable
length of time (Schofield: Teach Yourself Beekeeping p. 100).
crann mbech. A swarm of honeybees which has not been hived by a beekeeper
almost always makes its nest in a hole in a tree. The late Professor Greene has
drawn attention to the occurrence of the phrase crann beach in Muireadhach
Albanach’s elegy on his wife (Quiggin: Poems from the Book of the Dean of
Lismore xxvii §7). Here the poet compares the death of his wife to the robbing
of the honey from a bee’s nest: crann beach arna argain im fhail (cran bea 1 er
ni ergin malli MS) ‘ a tree of bees has been plundered beside me.’ The phrase
also occurs in H.3.18 p. 425b (our L — see Appendix 4 (c)) diri craind bech bi[d]
co .uii. saithi. But here crand [m]bech seems to be taken by the commentator to
mean ‘ a hive of bees ’ (cf. in cheis feiti) which would not suit the BB context.
glossa. The first glossator suggests that the tree in which the bees have made
their home is a bile or ‘ sacred tree ’ (often growing at tribal inauguration sites).
However, this narrowing-down of the application of crann is quite unjustified
by the text.
do fiur. These words have been omitted by the main scribe (cf. §43) and
added above the line {do fir ) in a hand which seems to be that of the second
glossator (see Introd. p. 5).
fod-gaib. As crann is neuter, we cut out the -n- of fodngaib A.
§48
The finder of a colony of bees in private property which is not faithche gets
one third of their produce (one can take trian to stand for trian a thoraid, cf. §46
cethramthain a thoraid). In old Swedish law the finder of a colony of bees in
privately owned forest gets the same proportion (Das Ostgotenrecht ed. Strauch
p. 217). However, the Swedish lawyer seems to have known less about bees than
his Irish counterpart, as he awards one third of the bees (not of the honey) to
the finder. Only in exceptional circumstances does a colony contain more than
one fertilized queen ( The World of the Honeybee p. 157). Hence a share of the
bees of a colony would be of no value unless there was a queen among them.
ruud. Here and at §49 we emend rud of the MS to ruud (ro ‘ great ’ + fid
(<C*widus) ‘ wood ’). The form ruud is unattested but Thurneysen suggests in
ZCPxiv 5 that the spelling rund in AL iv 278.8 (= Eriu xxii 82w) is for ruud with
-n- miswritten for -u-.

§48]
NOTES
157
The only occurrence of the word in a MS of the O. Ir. period is gen. sg. roida
{ T/ies . ii 290.11) from *roideo with depalatalization of the d through influence
of the nom. Heptad lxiv ( CIH 55.19 etc. = AL v 320) includes the frithe roida
0 raedha , roeda, raoda, rofida , rofeda v. 11.) as one of the seven estrays in Irish law.
Words of similar formation to ruud are roilbe <C ro + slebe (probably a
collective of sliab, see note to §30 lestrae) lit. ‘ a great expanse of mountain ’ and
romuir < ro + muir lit. ‘a great expanse of sea \ Already in O. Ir. these words
seem to have become synonymous with their bases fid , sliab and muir. Hence
the compiler of BB felt it necessary to supply the adjective mar ‘ great ’ after
ruud.
Down to the 17th century, large areas of the country were wooded; see G.
F. Mitchell, The Irish Landscape pp. 192-199.
ecmacht. Thurneysen ( Grammar §345) takes ecmacht ‘ impossible, hardly
possible ’, soch{u)macht ‘ possible, capable ’, dochumacht ‘ hardly possible ’ to
be compounds of the O. Ir. neuter io- stem cumachtae ‘ power ’ (a noun formed
from the vb. con-icc ‘ is able ’, vb. n. cumang). However, there are no certain
cases where a noun in -{d)e has lost its ending when compounded with so-, do-,
or a preposition. In his list of adjectives of this type ( Norsk Tidsskrift for
Sprogvidenskap xiii 340-3) Marstrander includes dothe{i)ng ‘ evil-tongued ’
« tengae) and sulabair ‘ eloquent ’ (< labrae ). However, dothe{i)ng seems to
be merely a scribal variant of the much commoner doithnge (also dothenga), cf.
soithnge ‘ well-spoken ’ (never *sotheing). The adjectives sulb{a)ir ‘ eloquent ’,
dulbair ‘ slow of speech ’ may be from labar ‘ talkative ’ rather than the noun
labrae ‘ utterance ’ (though it must be admitted that such adjectives are usually
formed from nouns rather than from another adjective).
We suggest that ecmacht, soch{u)macht and dochumacht are compounds of
an earlier *cumacht (cf. W. cyfoeth ‘ power ’) rather than O. Ir. cumachtae
(> Mid. Ir. cumachta). In the Irish Grammatical Tracts Deck §25 cumhacht is
given as a by-form of cumhachta, but this is perhaps a later development rather
than a survival of *cumacht postulated above.
In BB ecmacht is used as a noun in the meaning ‘ a place which is difficult
of access \ Both here and in §49 it is glossed .i. in lacha (cf. CIH 59.7 ecmacht
./. loch) but Plummer points out in Erin ix 40 that a lake is to be taken merely
as one case of an ecmacht : glossators often use to introduce an example
rather than an identification. Presumably in practice a person who found a
colony of bees in a cliff-face or in a tree growing in a marsh, lake-edge, lakeisland, river or cliff would claim that it was in an ecmacht, and so belonged to
him (§49 is diles do suidiu ).
dirainn. The literal meaning of dirann is ‘ that which is undivided ’, whence,
‘ land which has not been shared out.’ No doubt this term overlaps to some
extent with ruud and ecmacht. The A glossator takes it to refer to mountain (./'.
in sleibe) and it seems probable that even at the time of the text’s composition
most unshared land would have been mountain or moor. Thus the phrase
dir[a]ind uas each ‘ the unshared land above all ’ occurs in the O. Ir. text of the
laws {CIH 241.28; 394.33) implying that unshared land was generally at a higher
altitude than shared land. However, there is a reference in ‘ Privileges etc. of
Poets ’ to diorainn do muigh no sleibh no uisge ‘ unshared plain or mountain or
water ’ CIH 1125.16 {Eriu xiii 33.19; see note on p. 225). In O. Ir. commentary
to B Comaithchesa {CIH 198.4-6 = 579.6-8) dirann includes rudh 1 roilbe 7 foach
tuaithi ‘ forest and mountain and tribal wasteland (?).’

i58
NOTES
[§49
§49
ecmacht. The example given in glossc by the fourth glossator ./. locha be!
set is perhaps a lake in the Gaitee mountains; see Hogan’s Onomasticon and Dr.
Charles Bowen’s forthcoming ed. of the Dinnshenchas of Munster. There was
also a lake of the same name in Donegal ( Onomasticon ). Cf. CIH 49.15.
oenruidles. A number of ruidlesa tuaithe are discussed at the beginning of
Berrad Airechta { CIH 591.9 — 592.21, transl. Biirgschaft, § §1-20), but they are
all cases in which assets change hands, not exact parallels to the present case.
This is, however, only a consequence of the pre-occupation of that text with
such transactions and the sureties by which they are guaranteed. The nature of
the immunity is always determined by the context, e.g. CU §13 {Erin xvii 70),
Patric. Texts 174.7 ( ogdiles ).
acht cuit n-agi fine. For acht governing the acc. see note on CU §5 {Erin
xvii 79), V. Hull ‘ Old Irish acht ’, ZCP xxv 237-42. This usage is reliably attested
in the lawtracts. Though DIL s.v. acht puts a number of non-legal exx. under
the heading ‘ as prep, with acc.’ none of them clearly governs the accusative.
The claims of a man’s kindred to a share of his acquisitions are also referred
to in the tract Corns Bescnai , CIH 532.28-30; 533.171'.; 534.20-1. In the first two
passages it is a question of adnacal cona anrtechtaib imnai (532.1) ‘ burial with
its perquisites in regard to bequest ’. If a man was buried in the relec of a church
he was expected to make a bequest to that church. The value of the bequest was
limited by four considerations: the status of the testator, the balance between
his acquisitions and losses while in control of his inheritance, the extent to which
his inheritance was instrumental in securing his acquisitions, and the residual
claims of his kindred even if the acquisitions were not secured by means of his
inheritance. A third share seems to be the extent of the kindred’s due if the
inheritance is not instrumental in the acquisition, as in the case of bees being
found (533.19). In our text, however, we have to do with the cuit agi fine, not
the techta fine of CIH. 532.29. This, presumably, is the explanation of the trian
as each triun : one third is the share of the fine, one third of that third the share
which goes to the head of the kindred. His authority within the kindred is clearly
indicated (contrary to the argument of Thurneysen, ZCP xiv 369-70, that he
acquired this position after the period of the tracts). The B gloss1 is either a
misunderstanding or, more probably, represents a significant change in the law.
It appears to presuppose an initial division by which one third was the trian
flatha, one third the trian eclasa and one third the trian fine cf. CIH 910.27-8.
The notion that the man in question retains the trian fine is a complete misrepresentation of the text. The text itself, moreover, does not refer to a flaith at all,
only the age fine and the eclais.
cuit n-ecailse frisa mbi audacht. A’s eel- is presumably for Mid. Ir. eclaise,
a metathesised form of O. Ir. ecailse, (Ml. Wb.) alongside O. Ir. eco/so (Wb.).
Prim. Welsh *egl<>s was treated in Ir. as a fern, d-stem or /-stem. On the other
hand it may be that, at the period of composition, the svarabhakti vowels had
not yet appeared so that the forms were then *eclse and *eclso with syllabic /
(cf. Grammar §112).
Eclais appears to have two senses in this paragraph, not only ‘ church ’ but
also ‘ head of a church ’ (cf. flaith ‘ lord, ruler/lordship, realm ’ ; for exx. CIH
2211.3; 2262.19). Hence eclais here is the counterpart to the age fine, as the
sense requires. The head of the church is to receive a ninth share as will the head
of a kindred. In the last line of the paragraph, however, eclais is the counterpart

NOTES
159
[ §49
to fine, the whole social group belonging to a particular church, and has members
0 membur ) as does the fine.
For the meaning of audacht see F. Kelly (ed.), Audacht Morainn, p. 22. The
present text is important evidence for its significance. There are two main types
of evidence: some rather obscure legal material and the references in the
Additamenta in the Book of Armagh. These may be taken in turn.
It is not clear that the legal texts (as opposed to glosses) make any absolute
distinction between imnae ( timnae ) and audacht. CIH 529.23-4 lists six things
due from the tiiath to a church when the latter is in a proper spiritual and moral
condition : ubairt, dechmad, primite, primgeine, udacht, imna. After more general
material, this theme is taken up again at CIH 530.32 where dechmada, primite, and
primgeine are cited as dliged ecalsa dia memraib (probably the manaig). These
are then discussed in reverse order (531.3f., 16f., 20f.). With 532. If. begins
discussion of imnae which continues until 534.15, whereupon the tract turns to
a brief note on ubairt. After this it concerns itself with goire almost till the end
of the full text (534.26-536.15). Unless audacht has something to do with goire
or imnae it appears that it, alone of the six, has received no further treatment.
Undoubtedly the goire owed to, and sometimes by, churches was of importance,
but it is clear that the discussion of imnae is likely to have covered, by implication,
audacht as well. The glosses tend to distinguish the two (for ex. 529.34 ./. fri bas
for udacht in the text and ./. a nertslainte for imna), but the actual discussion of
imnae opens with the sentence (532.1): Cach n-adnacal cona urtechta imnai do
eclais caich iarna miad, ‘ every burial together with the perquisites of bequest
belongs to each person’s church in accordance with his rank ’. Imnae here
appears to be a donatio mortis causa though not necessarily a donatio ad mortem.
Similarly in the Heptad (incomplete), Biirgschaft p. 21 = CIH 596.29-30, we
have timnae fri hec as one of the seven things which defeat any dissolution of
ownership. It is, then, evident that the imnae may be as much fri hec as is the
audacht.
Two texts among the Additamenta in the Book of Armagh use the term audacht
( edocht , idacht ). The first ( Patric . Texts 172) determines the rules by which the
head of the monastery of Druim Lias is to be appointed. It is a coibse ( confessio ,
in the sense here of public declaration rather than private confidence) and an
edocht, made two years before Feith Fio’s death. Feith Fio appears to be the
erlam, founder of the monastery, and the text thus makes provision for the
succession after his death. It can, therefore, be said to be fri hec, in the sense
of mortis causa rather than ‘ at death ’. Furthermore, edocht here is not a
bequest of property but an ordinance determining what shall happen after the
death of the man making the ordinance.
The second ( Patric . Texts 178) describes provisions made by Bishop Aed of
Sleibte (Sletty, Co. Carlow) by which his church was subjected to Armagh. Aed
is said to have gone to Armagh and to have taken an edoct to Segene at Armagh
(Birt edoct cu Segene du Ardd Machae). Segene then returned Aed’s idacht to
him and Aed offered his idacht and his cenel and his eclis to Patrick till Doomsday. Aed is then said to have left his idacht with Conchad who went to Armagh
where Fland Feblae gave his church to him and ‘ accepted him also as abbot ’.
The following interpretation of the text is based upon that given by MacNeill
in his paper ‘ The Earliest Lives of St. Patrick ’ JRSAI Iviii, part i (1928) reprinted
in his Saint Patrick (Dublin and London, 1964). For further observations see
K. Hughes, The Church in Early Irish Society (London, 1966) pp. 86, 115-9.
Aed’s edocht or idacht is apparently a written text which can be taken to Armagh
and left there. It exists in more than one copy since one is offered to Patrick
till Doomsday and another is left with Conchad. The people involved are

i6o
NOTES
§49]
identifiable: Aed died in 699 = 700 ( AU ) but had probably already retired
from the bishopric of Sleibte since his obit describes him as anchorita. Segene
was bishop of Armagh from 661 to 688. It is likely that Conchad was Aed’s
successor as bishop of Sleibte; and, if so, he may well be the Conchad whose
obit as bishop is given in A V s.a. 691 = 692. Fland Feblae was Segene’s successor
as bishop of Armagh and died in 715. Sleibte is the chief church (primchell) of Ui
Bairrche, a dynasty of considerable importance in Leinster during the sixth and
seventh centuries (Byrne, Irish Kings and High-Kings, 131, 136-7). During the
lifetime of Aed the Ul Dunlainge of the Liffey plain had not yet finally won the
unbroken dominance in Leinster that they were to hold from the mid-eighth to
the eleventh century, but the Ui Bairrche were already threatened by the
aspirations of their northern neighbours, as they were by the Ui Felmeda branch
of Ui Chennselaig, their close neighbours to the south-east (cf. the supposed
burial of Crimthann Mac Endi Ceinnselich at Sleibte, Additamenta §14(3),
presumably a claim by Sleibte to the patronage of the Ui Chennselaig). The
claims of Kildare followed those of the rising Ui Dunlainge in the seventh century
(Byrne, 152), but a note appended to the Liber Ange/i in the Book of Armagh
( Patric . Texts, 190) shows that the heirs of Patrick had no intention of abandoning
their claims to authority in parts of Leinster. Kildare, of course, held an unchallengeable position in North Leinster. Aed then had every reason to turn to
Armagh: further away and enjoying little secure authority in Leinster, Armagh’s
demands would be far less onerous than the ecclesiastical wing of the grasping
Ui Dunlainge. Aed’s edocht was a defensive move. Admittedly it effected the
subjection of Sleibte to Armagh; but the text itself indicates that Aed could
secure the succession of Conchad who was subsequently accepted as abbot (not
the more lowly praepositus ) of the monastery, in addition to being bishop of
the see. It is important to note that Aed offered his cenel and his eclis together
with his idacht to Armagh. These were not, one may suppose, separate transactions. The idacht is likely to have contained provisions subjecting the kindred
and the church to Armagh, and perhaps also securing the continuance of the
abbacy in Aed’s kindred. On the translation of this passage see now F. J. Byrne,
Erin xxxiii 169.
The audacht, then, is a solemn declaration, occasionally written down, by a
man in authority which prescribes the conduct of affairs, within a sphere
considered subject to that authority, after his death or retirement. It is a device
which prolongs authority beyond its natural span and thus secures a continuity
of arrangements over a generation or more. The crucial point shown by the
Armagh texts is that the audacht could be used to determine the conduct of a
man’s kindred and even to subject it to a new authority external to the kindred
itself. One external authority to which a man — provided, we may suppose, he
was head of his kindred and had their agreement — might seek to subject his
kindred was that of a church (cf. CGH p. 90. f. 127a25: Ad-ropart Barr a huu 7
a hindhuu eter bin 1 marbu do epscup Ibar co brath). It may be in such a way that
churches acquired most of their manaig, monastic tenants. These manaig were
part of their churches and to the latter they owed the than n-imnai referred to in
the O. Ir. Riagail Patraic ( Erin i 219 §8). It is very probable that this trian
n-imnai is identical with the trian mentioned here from which the head of the
church took his third. The reference in this paragraph will, therefore, be to the
relationship between a church and a manach, a relationship created by an audacht.
Part of the relationship is that, in respect of acquisitions, his church stands to a
manach as a fine does to an ordinary layman. If, then, it is a manach who acquires
a frith mbech, his church is entitled to a third, and the head of the church to a
third of that third.

§§49, 50]
NOTES
161
arnacon derbarthar. Glossj arnara cain-diubarthar suggests that one should
read derba[r]thcir . There is a similar problem in B Crolige §47 where there is a
form na derbabar, glossed na rodiubarar , which Dr. Binchy emended to na
derbarthar. In both cases the sense is clearly ‘ defrauds, deprives unjustly and
this is precisely the meaning of do-opir (j di-uss-ber -). The reading of the MS can
only be defended if the verb do-rorban { di-ro-{ro-)ben -) can have the same sense
as do-opir. This is very doubtful: do-rorban means ‘ prevents, obstructs’. It
governs either the person prevented (e.g. ni derban each a chele , Thes. ii 294.1) or
the action prevented (vb. n. or clause). In this case one would have to interpret
arnacon derbathar eclais . . . di neoch as ‘ so that a church may not be hindered
from [receiving] anything ’. On the whole it is much more likely that derbarthar
is the original reading.
a membur. Since do-sli is used, e.g. CIH 533.19, in the context of acquiring
something, it is probably correct to construe membur as nom. pi. neut. as in the
Cambrai Homily {Thes. ii 246.6). It would also be possible to take a membur
as the prep, a ‘ from, out of’ with dat. sg. membur. In either case the clause
refers to the share due to church or kindred from the acquisitions of their
members.
§50
The last part of BB, apart from the colophon, deals with theft of bees.
Interpretation is made more difficult because the tract on theft, Bretha im Gait a
(BG) is incomplete (ed. V. Hull, ZCP xxv 211-25). Yet this makes the evidence
of BB all the more precious, indeed indispensable to an account of the law of
theft. The penalty for theft is divided into two parts, aithgein and log n-enech.
Aithgein is due to the owner of the stolen article, the set in the terminology of
BG and BB (§51). The owner of the set receives a sum which is the value of his
set multiplied by a figure which varies according to the nature of the set {CIH
478.8-10). The owner of the place from which the set was stolen receives the
whole or a part of his honour-price according to the nature of the place— for
example, if the set was stolen from a man’s house he is entitled to his full honourprice {CIH 478.21). The owner of the set and the owner of the place are not
necessarily the same person. If the set of a king’s neighbour is stolen from the
house of a king, the full honour-price of the king is due. Of this, however, the
king keeps only two thirds; one third goes to the owner of the set who also, of
course, receives aithgein {CIH 478.26-7). The places from which a set may be
stolen are divided into three categories: (1) tech or treb (cf. BB §50 amal bid a
treib rosn-uccad; §51 it comdiri fri seotu trebe ); (2) faithche {BB §52) or airlise
{CG 218, cf. 123 for tech); (3) sechtar faithchi {BB §53). The evidence of CG 218
suggests that half the owner’s honour-price was due if it was the airlise or faithche
from which a set was stolen. The fraction due if it was sechtar faithchi is not
known. The glosses 53e, first and fourth glossators, put the rate at one seventh
of the honour-price. They may be right, though a third and a quarter are also
familiar fractions of an honour-price (e.g. CIH 437.13-4, 19-20, 24, 26-7).
BB' s use of the term dire is far from clear. The text is concerned to do
two things: (1) establish a rate of dire for bees which should be the same as the
rate for certain other possessions; (2) assert that the full rate of dire should be
paid (what that full rate is depends upon the circumstances). To make the first
point BB uses the phrases rosuidigthea i comdiriu fri (§50; similarly §§52, 53)
and it comdiri fri (§51). To make the second it uses the phras t di-renar landire
(§§52, 53). The trouble is that dire can mean either ‘ penalty ’ or, more specifically
‘ honour-price ’. Landire in §§52 and 53 cannot simply mean ‘ full honour-price ’
since the whole point of the distinction between treb, faithche and sechtar

162
NOTES
[ §50
faithchi , in the context of theft, is that full honour-price is due for theft from a
treb, but diminishing fractions of a person’s honour-price are due if the theft is
from a faithche or sechtar faithchi. The Idndire of §§52 and 53 cannot, therefore,
mean ‘full honour-price’ since those §§ refer to the thefts from faithche and
sechtar faithchi , not to thefts from a treb. Yet, if Idndire cannot mean ‘ full
honour-price’, it cannot refer to a full restitution ( aithgein ) either: not only
would it be highly confusing for dire to refer to aithgein, but the distinction
between treb, faithche and sechtar faithchi is relevant to the issue of honour-price
and not to that of restitution. For honour-price one asks ‘ From where has it
been stolen?’; but for restitution the question is ‘ What has been stolen?’ It
seems, therefore, that Idndire must refer to honour-price. The only possible
interpretation is that Idndire means ‘ honour-price at the full rate of honour-price
appropriate to treb or faithche or sechtar faithchi, whichever it may be.’ This
may seem over-elaborate, but it only shows again that the law of bees is being
grafted onto the stock provided by old and fundamental areas of Irish law.
It may still be thought strange that bees are equated with seoit trebe in §51,
with uasalnemid chethrae in §52 and with lu-chethrae in §53. Set, as has been
pointed out above, can simply be a general term for an article of greater or less
value, in this context a stolen article. There is, therefore, nothing odd about the
equation with seoit trebe. What remains to be explained is why bees are equated
with uasalnemid chethrae in §52 but lu-chethrae in §53. A clue to the point of
difference is provided by a rule in BG ( CIH 478.21-2): Ni gatar a treb direnar
lanlog enech caich asa treb(a ) acht ni etirscara lu de, ‘ Anything stolen from a
homestead is paid for at the full honour-price of him whose homestead it is
except for anything which low value {lu) remits from it.’ According to the gloss,
if an article is of low value, half honour-price is due. This may be right, but it
cannot be relied upon especially as the phrase ni etirscara lu de ‘ anything which
low value remits from it ’ does not suggest a single standard reduction. In the
case of lu-chethrae, however, there must have been a standard reduction, otherwise
they would not provide a standard to which bees could be equated. The point
of equating bees within the faithche with nemid chethrae, and bees sechtar faithchi
with lu-chethrae, can now be examined. What it does is to reduce the proportion
of the honour-price relatively quickly for bees sechtar faithchi compared with
bees within a faithche. Whatever the rate of reduction for other animals may have
been, the reduction for bees will have been greater. A large and valuable animal
such as a milch-cow will still have been a nemed cethrae even if sechtar faithchi.
Ordinary animals, in other words, will have been intrinsically either nemid
chethrae or lu-chethrae. Bees, because they are a colony or hive rather than
individuals for legal purposes, fit neither category obviously and so could be
shifted from one to the other. It appears that airlise in CG 218ff. plays a part in
the law of theft equivalent to faithche in BB. It may also be true — though this
is less certain — that the general term fonts in CG 220 refers back to airlise in the
previous sentence. If these two suggestions are correct then the scheme for
ordinary animals was as follows (values dependent on the unreliable evidence
of glosses are in square brackets) :—
treb I tech
faithche
sechtar faithchi
Log n-enech/dire
ordinary possessions
(including
nemid chethrae)
full honour-price
b honour-price
i honour-price
small possessions
(including lu-chethrae)
[|- honour-price]
[i honour-price]
[£ honour-price]

§50]
NOTES
163
If this scheme is correct the corresponding scheme for bees has two notable
differences: —
bees within les or lubgort : full honour-price
bees within the faithche : \ honour-price
bees sechtar faithchi : [| honour-price]
First, treb includes everything within les and lubgort. The inclusion of the les
is unremarkable: CG 210 assigns the same honour-price to entering a les
without permission and opening the house without permission; so treb in BB
might include everything within the les, even though CG talks only of tech as
opposed to airlise in the context of theft (11. 211-9) A lubgort, however, may
often have been sited outside, though adjacent to, the les. That it is covered by
the full honour-price suggests that it was standard practice to place a hive in a
vegetable garden. By contrast the small fraction of honour-price due to the
owner of the land for theft of bees sechtar faithchi suggests that it was not usual
for bee-keepers to move their hives far from the house in order to find good
grazing, or at least that the author of BB did not favour the practice. The value
of one eighth of the honour-price — derived by halving the quarter due for ordinary
possessions — is pretty close to the seventh of the glossator, §53 e, but in general
the glossators seem to have been at sea partly because they used the term rnaigen
or maigen digona meaning ‘ precinct ’ and were unsure how to relate maigen to
faithche (§§52a, 53e; cf. Appendix 1 (h), Appendix 2 (k)). The interest of this
section of BB, then, stems from the way in which the rules for theft are modified
to suit the particular case of bees: the author of BB adheres to the general law
of theft, as he does for comaithches, but he does not allow his adherence to
prevent him from stating special rules for bees which are not in strict agreement
with the ordinary law.
The final point of interest about this part of BB is not something which it
states, but something which it omits. BB makes no mention of aithgein for theft
of bees, only of dire (namely log n-enech). Yet aithgein was the main compensation for the normal owner of an animal. Its importance is much reduced if
the owner of the animal is a king or other nemed, but to a freeman such as the
ordinary boaire with an honour-price of five sets aithgein is vital. If the owner of
the land from which something is stolen is a king or other nemed, the owner of
the stolen article receives a third of the king’s or nemed' s honour-price leaving
two thirds to the king or nemed ( CIH 478.26-9). If the owner of the stolen article
was a king or other nemed, but the owner of the land from which it was stolen
was of lower status, the thief must pay separately the honour-price of the king or
other nemed to the king or nemed and the honour-price of the non -nemed to the
non-nemed (CIH 479.6-9). To see what all this implies let us assume that one cow,
worth two sets, has been stolen, that the king has an honour-price of seven cumals,
that the other party is a boaire with an honour-price of five sets and that the
cumal has the value of five cows (one of several values given it in the texts): —
(1) the owner of the land is a king; the owner of the stolen cow is a boaire : — -
king gets (rounding up) 24 cows from log n-enech;
boaire gets (rounding down) 1 1 cows from log n-enech ;
boaire gets 5 cows from aithgein, i.e. a total of 16 cows.
Total : 40 cows
(2) the owner of the land is a boaire, the owner of the cow is a king:—
the king gets 35 cows as his log n-enech;
the king gets 5 cows as his aithgein;
Total: 40 cows;
the boaire gets 2\ cows (= 5 heifers) as his log n-enech;
Total: 42i cows.

164
NOTES
[ §50
The arrangement clearly favours the king; but whereas aithgein is only of the
total received by the king in the second case, it is nearly a half of the total received
by the boaire in the first. If the owner of the land had been, not a king, but
another boaire, the aithgein would form the greater part of the sum received by
the first boaire (5 cows as against, probably, 2\ cows).
We have no reason to suppose that ownership of bees was restricted to kings,
and nemid such as high ecclesiastics. The absence of aithgein from BB, therefore,
is a puzzle. The solution may be that a bee-hive was unlikely to be moved from
its owner’s land unless it were sold or given away. The reason why the ordinary
law of theft is interested in valuables stolen from land belonging to a third party
may be that ‘joint-herding’, comingaire, made this a very common situation.
A short tract makes special provision for the pasturing of cattle, pigs and sheep
in a joint herd which spent a certain period on the land of each participant ( CIH
576.24 — 577.24). If this practice was at all common, the law of theft would be
bound to pay particular attention to valuables stolen from a third party’s land,
quite apart from the widespread use of pledges to guarantee obligations. Beehives, however, may hardly ever have been sited on another’s land. It may not
have been common for bee-hives to be moved to another place for the sake of
special grazing (though see Introd. p. 45). In this case the whole compensation
for bees may have consisted of the owner’s honour-price or a fraction of it. Bees
may have been considered as lu-chethrae by nature, so that their elevation to the
status of nemid chethrae within the faithche will have counteracted the effect of
the absence of aithgein. Furthermore, if the bees kept in seventh-century Ireland
produced a large number of swarms, the supply of bees may have tended to exceed
the demand, in which case the value of a hive, and thus of its aithgein, may have
been relatively low. (But cf. the values of Welsh hives and swarms tabulated in
Appendix 7, pp. 204).
It should also be noted that the whole principle of the scheme was stated
very well in §54: Di-ren-sede fo choibni co tabarr techte do magin dia foxlaiter,
‘ he pays according to “ kinship ” so that the due is given to [the owner of] the
open land from which they are removed ’. Coibne, ‘ kinship ’, is the metaphor
used in BG ( CIH 479.8) for the equality of the honour-price due to persons of
the same rank for theft from their houses.
i llugburt no i llius. Cf. Patric. Texts 174.7 (= Thes. ii 239.16) cona llius i a
llubgort. We emend llugbart of the MS to llugburt. We retain the metathesis,
as lub-gort ‘ garden * is spelt with b and g transposed in O. Ir. texts e.g. lugbart
Ml. 1 21cl 2, hi lugburt Thes. ii 294.16. In §51 we emend luburt to lugburt.
foda-rotlae no roda-gatta. The use of the verbs for theft in BB is interesting.
Here fo-tlen is paired with gataid; in §52 fo-tlen is paired with fo-coislea; in §53
fo-coislea with gataid. In §54 teol and taide are used to define the meaning of
fo-coislea more closely. Gataid is a general word for ‘ steals ’, but the others are
capable of more limited meanings. Foxal, the vb. n. of fo-coislea, may contrast
with tothlae, vb. n. of do-tlen, as in Cain Lanamna §4 (Stud. E. Ir. Law p. 16).
Lanamnas foxail is a form of union in which a man removes a woman against
the will of her kin. It is normally public. Lanamnas tothlai, also termed lanamnas
taide or lanamnas tothlai i taidi, is a union in which the man secretly visits the
woman. There is a contrast of direction and a contrast between publicity and
secrecy. Fo-coislea, however, is not necessarily public: e.g. Ni fo-roxla taide no
cnet no fogurred no forcomul 11-ecne as-renar . . . (Cain Lanamna §18) where the
removal is first secret, then attended with groans, then with loud protest and
finally is forcible. Fo-coislea then is capable of contrasting with compounds of

§§50, 52]
NOTES
165
tlenaid and with taide, but need not do so. Among the compounds of tlenaid ,
fo-tlen has, it seems, no specific directional sense, though in the nature of things
it is frequently used of taking something away; in CIH 578.1 Of. there is an overingenious distinction between tothlae and fothlae which presupposes that in
both cases something is brought onto a particular piece of land. Tothlae and
dithle have, however, opposite directional sense: tothlae is bringing something
in, dithle taking something away. In both cases it is usually secret, and thus
unlawful, but need not be (e.g. Celtica ix 158, 1. 50). On teol (to tlenaid ) see
Thurneysen, note ZCP xx 212. In BB §50, therefore, there may be a distinction
between foda-rothlae and roda-gatta in that the former refers specifically to secret
theft; and there may be a similar distinction in §52 between foda-rothlae and
foda-roxla; but in §53 there can hardly be any distinction. In §54 the reference
to secrecy is explicit, but only through use of the phrase fo theol -j taidi (for this
phrase, with similar glosses, cf. CIH 2336.14).
For the glosses on the various verbs for theft in §§50-53, see Introd. p. 20.
diren-side. We emend dirense to diren-side on the basis of the gloss eirnid
seisideic , as side is regularly glossed by seiside , seisideic , seisideig, sideic ; cf. §2
dlid-side gl. airiltnigid seisidhe , §12 alid-side gl. airiltnigidh seisidheig.
DIL s.v. 4 se suggests that dirense contains a 3 sg. masc. emphatic pronoun
se (?), but this suggestion is not supported by the gloss. The other example
given in DIL is not genuine as it derives from a wrong word-division by the editors
of AL iv 202.3 (= BB §54). Here the MS has dirensede (gl. eirnid side) where
sede is a variant spelling of side, cf. ni farnic sede Wb. 2a21. The AL editors
divide diren se de which makes no sense.
§52
fri huasalneimthiu cethrae. We read acc. pi. -neimthiu with A, cf. nemthiu,
LU 9786; neimte « -hi), CIH 580.37. B has -nemeth, which is probably acc. sg.
but might possibly be acc. pi. neuter (it is glossed by plural nouns ./. lulachacha
1 daim riata ). However, when used of persons or animals nemed seems always
to be masculine (see notes to §15).
It is clear from the context that liasalnemid chethrae here refers to the larger
and more valuable domestic animals, while lii-chethrae in §53 refers to the smaller
ones. This two-fold division of domestic animals is not found elsewhere in the
text of the laws, but is quite common in later commentary, (e.g. CIH 833.22;
864.16 morcethra, mincethra). The term nemid chethrae is attested in CIH 38.19
(cf. 897.11), but is used with a more specialised meaning of livestock which are
temporarily or permanently immune from distraint (see note to §15 nemed (4)).
The term lu-chethrae is not attested elsewhere.
The B glossator gives ‘ milch cows and trained oxen ’ as examples of
liasalnemid chethrae and ‘ little lambs, little pigs, little calves and kids ’ as
examples of lu-chethrae. However, this classification omits full-grown sheep,
pigs and goats. It seems probable that the compiler of BB intended the liasalnemid
chethrae to include cows, oxen, bulls and horses, and the lu-chethrae to include
sheep, goats, pigs and the young of the cow and horse. One can compare the
distinction between large and small animals in Welsh law {ALIW ii 708). Here
the sow, goat, sheep and calf are given as examples of small animals (anifeil
by chan).
A glossators II, III, and IV attempt to explain §§52-3 with the help of §2
of Bretha im Gatta ‘ judgments about thefts Like BB, BG deals with the theft
of domestic animals. However, it does not mention the theft of bees, and does

NOTES
[§§52, 53
166
not make BB' s distinction between larger and smaller domestic animals. Its
regulations concerning the restitution of stolen animals are from Exodus xx ii (see
note to 52d) whereas BB contains nothing of Biblical origin.
glosses 52b> c. See Introd. p. 20.
gloss 52d III. The third glossator’s a ceathair i men ‘ four for one ’ is a
quotation from BG. This fragmentary text has been edited by Hull in ZCP
xxv 211-25. The main source is H.2.15A p. 38, but a few quotations are found
in other MSS. §2 of BG deals with restitution for the theft of domestic animals,
and is quoted by BB glossators because of its supposed relevance to the theft of
bees. Hull points out in a note on p. 218 that the same rate of restitution is found
in An Irish Penitential (ed. Gwynn, Erin vii 154 §3) and in Die irische Kanonensammlung Lib. xxxix cap. 3 (ed. Wasserschleben p. 99) and has its source in
Exodus xxii 1 .
The H.2.15A text has diabul naithgena i marbdile 7 i neochu 7 mucca ce tarda
i churcha cethuir i noen cuicthe i naithgin i mbu 7 i dumu ( CIH 478.8). The only
other text (H.3.17 col. 487A) has dublad naigena a marbdil- 7 a neacha 7 a muc-.
cethardha a caorcha ceathair a naon cuic- a naithgin i mbu 7 damh- ( CIH 1977.4).
Hull reads diabul n-aithgena i marbdile 7 i n-eochu 7 mucca; cetharda i curcha; [a]
cethuir i n-oencuiced i naithgin i mbu 7 i dumu which he translates ‘ a double
share of the restitution for inanimate chattels and for horses and pigs; a fourfold
amount for sheep; four(-fold) for the one-fifth [that constitutes the restitution
is] the restitution for cows and for oxen.’
Our gloss a ceathair i naen shows that Hull is right in his restoration [a]
cethuir, but suggests that he is wrong in taking oen to be compounded with the
following cuicthe (which he emends to cuiced). We put forward the following
reconstructed text and translation: diabul n-aithgeno i marbdile 7 i n-echu 7 i
mucca , cethardae i caircha, a cethair i n-oen, coicde i n-aithgin i mbuu 7 i ndamu
‘ double restitution for inanimate objects and for horses and for pigs, fourfold
[restitution] for sheep — four for one— fivefold in restitution for cows and oxen.’
We take cuicthe H.2.15A, cuic- H.3.17 to be for coicde 6 five things ’, Grammar
§387. It must be admitted that a cethair i n-oen seems to be superfluous, and may
possibly originate as an early gloss which became incorporated in the text of BG.
gloss 52d IV. The fourth glossator adds ceth-e a ncen here and cethr-e a men
at 53d. The abbreviation ceth{r)-e usually stands for cethraim(th)e (<0. Ir.
cethramthu ) ‘ a quarter ’, but as this does not suit the context we suggest that it
was intended for cetharda(e) ‘ four things ’, cf. im cetharda II.
§53
fri lu-chethrai. Following the reading of B, we take lu to be an adjectival
prefix, and cethrai to be acc. sg. of cethrae ‘ animals, livestock ’ (it could equally
be acc. pi.). Alternatively, one could read lii cethrae lit. ‘ a small one of livestock ’,
with lii used substantively and cethrae as gen. sg. For the meaning of lii-chethrae
see note on §52 fri huasalneimthiu cethrae.
gloss 53°. The fourth glossator’s addition to this gloss presents problems.
We read cairig, a fri indib7 sechtmad neneclainni ‘ i.e. sheep, three for them and
one seventh honour-price’ taking this to mean that the penalty for stealing a
sheep was to restore three sheep to the owner and pay him one seventh of his

§§53, 54J
NOTES
167
honour-price. However, as we have seen in the note to gloss 52d, BG states that
four sheep are to be restored for the theft of one. It is therefore possible that we
should here read leth rather than tri, taking it to be an / with a stroke through it
rather than a t with an i over it. The glossator may be expanding the first
glossator’s statement that the theft of small animals in a precinct entails half
honour-price ( letheineclann ). One can compare 52e where the fourth glossator
says that the theft of large animals in a precinct entails full honour-price (./. Ian
intib a maigiri).
§54
teol •] taide. For different types of theft, see notes to §50.
ro-ch-fintar. For the enclitic connective -ch, see Introd. p. 13.
ni imdich. DIL gives the deuterotonic form of this verb as imm-dich. The
long i is frequent in later manuscripts, but may not have been regular in O. Ir.
This verb is from *imb-di-wik- (lit. ‘ around-from-fights ’ i.e. protects) and is
thus a further compound of do-fich ( di-wik -) ‘ punishes, avenges ’.
The treatment of *di-wik - is variable in the O. Ir. Glosses. Thus in the 3 sg.
pres. ind. -dig (= -dich) Ml. 24b 17, the original disyllabic i-i has coalesced to
form long / (Thurneysen points out that such contractions in fully stressed
syllables are still rare in O. Ir., Grammar §114). In nimdichimse form naimtea
Ml. 38c21 (sic leg. Corrigenda Thes. i 719, translating ‘ I do not avenge myself
on my enemies ’) -dichim seems to be 1 sg. pres. ind. with original i-i contracted
to short i. Alternatively, one might read dichim on the model of -dig.
The disyllable seems to be preserved in 2 sg. ipv. deich Ml. 72dl 1 (for dei'ch.
Grammar §114) unless it is merely a variant of dich , cf. dofeich Wb. 6al6 for
regular dofich.
Where di-wik- has suffered vowel affection the resultant combination seems
always to be a disyllable (ie or id) rather than a diphthong ( ia ) in O. Ir. Thus the
manuscript has no lengthmark over the i in diagar Ml. 101cl6, diachtae 111 b6,
diachti 23dl8, diachtid 19dl5. The disyllabic id in the latter is confirmed by the
spelling diechtaid Ml. 83c2, as ie is very rarely written for the diphthong ia in
O. Ir. (Grammar §53).
To conclude, it is likely that at the time of BB’s composition the forms of
the cpd. *imb-cli-wik were 3 sg. pres. ind. deuterotonic *imb-diich, prototonic
-imdich. For the reduction of disyllabic ii to short i in unstressed syllable, cf.
3 sg. cons. pres, of the substantive vb. biidi> copula bid.
fo choibni. The basic meaning of coibne [com 4- fine ] is ‘ kinship, consanguinity ’ but it is also used to mean ‘ equivalence, appropriateness ’ (see DIL
s.v.). See end of first note to §50.
tabarr. We take tabar of the MS to be for original tabarr, prototonic 3 sg.
pres. ind. pass, of do-beir. Singular pres. ind. passives of this and other compounds of beirid are usually spelt with single -r from the Milan Glosses onwards,
e.g. dober Ml. 56bll (doberr Wb. 3cl), -tabar Ml. 52a2 (tabarr Wb. 17a3), eiper
CG 1 (-eperr Sg. 73all).
do tnagin. The sense requires the emendation of di MS to do. This is supported
by the gloss dir ind eraind ‘ to the owner of the land ’ where dir — do fir, cf. 23b,
37c, 46d, 48e.

1 68
NOTES
[§§54, 55
dia foxlaiter. We tentatively emend diandafoxlaiter MS to dia foxlaiter and
translate ‘ from which they (the bees) are removed The intrusive (n)da may
possibly have come in through confusion between fo-coislea [fo-com-sel-] and
do-foxla [ to-fo-com-sel -] of the same meaning. The spelling of the preverb doas da- is frequent in later MSS, and is particularly common in the glosses in A,
e.g. 54e dafintar, 54h dabiad.
Another possible explanation of this intrusive (n)da is that it arose through
contamination with foda-rothlae, foda-roxla, roda-gata, foda-coislea of the
preceding sentences, in all of which da is 3 pi. infixed pronoun.
§55
The same formula is used to close Coibnes Uisci Thairidne ( Erin xvii 72) : is
for sund ro suidiged coibnius uisce tairidne la Feniu and to close Bretha im Fuillema
Gel/ ( CIH 477.23 = AL v 422): is for sunn ro suidiged irnchomus gell la Feine.
The lost tract Fidbretha probably opened with the similar formula can
rosuidighthea fidbretha la- which is quoted in CIH 2106.3. See note on §14
a besaib fidbreth.
ADDENDUM
Note to §8 ind n-oircne. Dr. Binchy points out that Thurneysen's interpretation of ind n-oircne is supported by the gloss ind eirci i n-ind fogla the
minimum of penalty for the minimum of injury ’) on ind i n-araill eiranar, CIH
336.30-1 = 2215.29. The latter phrase is from a passage on the fines for various
degrees of personal injury ( CIH 336.20 ff. = 1266.1 ff.). The least serious
injury (ind fogla of the gloss) is the banbeim ‘ white blow ’, i.e. a blow which
causes no swelling or bruising.