Contributions to the History of the Deponent Verb in Irish. by J

444

XV1.-CONTRIBUTIONS TO THE HISTORY OF THE DEPONENT VERB IN IRISH. BY J. STRACHAN.

[Read at the Meeting of the Philological Society held on Friday, June lrf, 1894.1

THE object of this paper is not to investigate the origin of the T deponent, which Old Irish ahares with Latin, and its relation to the Indo-Germanic verbal system. or to discuss, except inci- dentally so far as they have any bearing on the subject proper, the theories that have been put forward concerning the origin of these forms.' Whether it will ever be possible to get beyond conflicting theories, and to arrive at any certain or even probable account of the genesis of the type, may be reasonably doubted. But, taking the deppnent as it exists in the oldest records of the Irish tongue, it should not be an impossible task to trace, with more or less exactness, its history within the Irish language itself, to follow the old forms in their life and decay, and to search out the starting-point and follow the development of any new types. The degree of precision with which such an investigation can be carried out must depend on the nature of the documents on which it is based. Where there is a continuous series of dated documents, each of which representa faithfully the language of its time, the course of the enquiry will run smoothly enough. In Irish, however, the student does not find himself in this fortunate position. For Old Irish we have trustworthy documents in the Glosses and in fragments of Irish preserved in the oldest manuscripts. In the later language we have metrical compositions like the Saltair na Ram, where the original forms were to a great extent protected by the metre. But with the mass of Middle Irish texts, some of them very ancient in their origin, preserved

1 Cf. Wmdisch, Ue6er die Perbdformen mit dem charakter r., etc., Abhandl. der Sachs. Ge6. d. miss., Phil.-hist. C1. x. 447 sq. ; Zimmer, Ueber das italo- keltische Passivum und Deponem, Kuhn's Zeitschrift, x5. 224 sq. ; Thurneysm. EZ xxxi. 62 sq., Indo-German. Forsch. i. 460 sq. ; Brugmann, Grmdriss, ii. 13886q. THE DEPONENT VERB IN IRISH-J. STKACHAN. 445 in manuscripts of the eleventh century and later, it is very different. In the transmission of such texts there was a tendency to replace old forms by modern ones. The probable extent of such corruption is not to be estimated d priori by the date of the manuscript. We shall find instances in which later manuscripts preserve the old forms better than earlier ones. And in the same manuscript different texts will be found to have suffered in a different degree. Thus, in the Book of Leinster, a manuscript of the twelfth century, the old deponential forms are well preserved in many of the old texts, but, for instance, in the famous 2"ia 36 Chailige, the Cattle-raid of Cooley, they have almost vanished. In such a case as this we must suppose that in the one instance the old forms were more or less faithfully copied, in the other the language has been approximated to the language of a later period that the old tale might delight new generations. Such corruptions of necessity make the problem more com- plicated. Already in the Glosses the deponent verb has begun to pass into the active; and in a particular active form from a deponent verb found in one of these old texts, preserved in a manuscript of a much later date, it may be impossible to say whether the active form was original or whether it was introduced later. The remedy here lies in widening the field of observation j for by the examination of a number of texts of the same character individual peculiarities may to a great extent be eliminated. Another difficulty is the difTiculty of chronology. In some cases the date of the composition of a text may be accurately or approximately fixed from the internal evidence supplied by the subject-matter. In other instances it might seem to be fixed by the name of the author. Unfortunately we must be very careful in accepting mch statements ; there is, throughout Irish literature, too much of a tendency to fasten to the productions of later times the great names of the past. A third means of fixing approximately the date of composition is the language. But here again the development of the Irish language in its successive stages is a subject that has yet to be worked out. Until that is done we must rely in this matter, to a great extent, on general impressions, which accurate observation may prove to be false. The following pages may serve as a starting-point for the history of one particular form in Irish, the deponent verb. The results of such an adventure on a fresh field must of necessity contain much that will have to be modified or 446 THE DEPONENT VERB IN IRISH-J. STR.4CHAN.

corrected by further observation and discussion. Something will have been accomplished if the way has been made smoother for those who come after. The main subject falls, naturally, into two parts. The first part contains a collection of materials for the history of the deponent verb; the second treats of the history of the deponent based on these materials; a third part will deal with some new Irish developments, whereby certain forms of the active verb have taken to themselves endings borrowed from the deponent inflexion. Part I., though it will consist chiefly of dry lists of forms, is a necessary preliminary to Part 11. For the Old Irish of the Glosses the Grammatica Celtica furnishes extensive 001- lections of material, though the further publication of Irish Glosses, particularly of the Nilan Glosses, enables considerable additions to be made to the material gathered together by Zeuss and Ebel. For the Old Irish of the profane literature and for Middle Irish there are no collections of any extent; here the work had practically to be done from the beginning. To ransack even as much of this literature as is already generally accessible would be an endless and a profitless task. What has been done has been to take a number of representative texts and to examine them carefully. This may lead to the omission of some deponent forms that might have found their place here; but, on the other hand, the history of the deponent is not to be learned from a multitude of scattered forms huddled together from a variety of heterogeneous sources. Apart from the glosses the material has been got to a great extent from the two oldest Middle Irish manuscripts, the Lea6har na TiUidhm (eleventh century) and the Book of Leinster (twelfth century). The former has been examined throughout; of the latter I have read through all that seemed to be of importance for this investigation. From these two sources have been selected such texts as appeared most illustrative. To these have been added texts from other sources, such as the Sanas Cormaic and the Saltak na Rann, two works which we shall find to be of the highest consequence, since between them lies the breaking up of the deponent inflexion. In the arrangement of the material collected from these texts that order has been adopted which seemed to put the facts in the clearest light; it does not of necessity correspond to the order of the date of the original composition of the several texts. In the present state of our knowledge it is impossible to arrange THE DEPONENT VERB 1N IRISH-J. STRACHAN. 447 the oldest stratum of texts chronologically, and something besides has been sacrificed to convenience. The language of the Old Irish Glosses has been treated as a whole. Each of the other texts has, as a rule, been treated separately. This involves repe- tition, but that is a much less evil than would have resulted from the mixing up even of the material got from closely similar texts. We have seen the corruptions to which these documents hare been exposed in the course of their transmission. But they did not all suffer in an equal degree. For example, to put together the forms of the deponent verbs found in the LU. and LL. versions of the T&n would be to convey an entirely false impression. Beeides, it will be found that the presentation of such forms from a number of texts of various kinds and of various ages will be the most effectual way of illustrating their history.

I. MATERIALS FOR THE HISTORY OF THE DEPONENT VERB.

1. THE OLD IRISHGLOSSES.' The deponent verbs here arrange themselves naturally in two classes. The first class consists partly of primary verbs like sechw, Lat. 8epuor, *tluohor, Lat. lopuor ; partly of denominative rerbs such as *comcalnw ' T fill,' from eomllira ' full ' ; cf. Latin denominatives in -8re. With this class goes a small number of rerbs that show deponent forms only in certain parts, generally in the perfect. Only those parts of the verb have been given in which the deponent inflexion differs from the active. That it map be clear at a glance how far the actire infiexion has

The references are to the following editions :- Acr. = Codex Auy~tmCarolisruhensis, ed. Stokes. Bcr. =Codex Be tx Caroksruhensis, ed. Stokes. Pcr. =Codex Prisciani Carolisruhensis, ed. Stokes. Wb. = Codex Paulinus Wirziburgensis, ed. Stokes. Ph.=Glosees on Philargyrius' Scholia on the Bucolics, ed. Stokes (KZ. xxxiii. 62 8 Trans. Phil. SOC.1893). Ml. =Codex Ami;osianus, ed. Ascoli. Sg. =Codex Sangallensis, ed. Ascoli. Tur. = Codex Taurinenais, ed. Zimmer. Cod. Cam. =Codex Camaracensis, ed. Zimmer. Bv. =Codex Bedse Vindobonensis, ed. Zimmer. Incant. Sg. =Incantationes Sangallenses, ed. Zimmer. Sp. =Camha Monasterii S. Pauli, ed. Windisch. 448 THE DEPONENT VERB IX IRISH-J. STRACHAN. encroached upon the deponent, the deponent and the active forms have been arranged in parallel columns.

DEPONENTFOEMS. ACTIVBFORMS. Bp,' timeo, ad-dggzcr. Pres. ind* ni dgor, non timeo, sg. 112." sg. 1. f nadnhggzcma, g. neminem me timere, MI, 74b 19. sg. 3. ni agathar, Wb. 18 3. nisnagathar, Wb. 68 7. inna agathar ni, annon timet aliquid, Ml. 87d 15. niconagathw, MI. 1298 12. inti! adagadar, qui timet, MI. 5314. nadnagathar, g. non timere (subj.?), Ml. 74b 21. nadagether, qui non timet, MI. 1298 2. pl. 3. ni ugetar, g. non verentur, hll. 3gb 14. subj. sg. 2. na digder, Sp. 111.8 2.2 3. corzlagathar, ut timuerit, Ml. 668 2. pl. 3. ugatar, MI. 51d 11. pret. pl. 3, conaruaigsetar, M1. 35~4. adraicheetar, Nl. 80d 4. adraigsetar, Ml. 124b 6 (bis). fut. sg. 1. adaichferea, g. timebo, Ml. 6BC 17.

1 In some of the forms of thi6 verb a transition is found to the i conjugation. The same variation is Been in other deponents, as comalnur (where it is employed to distinguish the passive from the deponent) and Zrimur. A similar phenomenon is seen in other cae.es. Thus from &dim, I say (=*r%heiC), come radas, M1. 42c 4, 10; niradut, 31a 18. Perhaps this confusion waa helped by cases like *frecruim, freerat (by the side of gairim, guret) =*frith-garim, *frith-garet, where on the loss of u the timbre of r was assimilated to that of the preceding consonant, this being expressed graphically by writing u after it ; cf. Brugmann, Grundriss, i. 4 640, note. So in the preterite of immrddiu, I think, imm-ro- rcidius gave regularly immrordus, a form like rocharus from caraim. 8 Cf. Thurneymn, 1%. Forsch. i. 460. THE DEPONENT VERB IN IRISH-J. STRACHAN. 449

DBPONENTFORMS. ACTIYB FORMS. 3. adaichfedar, qui timebit, M1. 46c 20. pl. 3. ni aichfetar, M1. 80b 13. frith+alim, praestulor. pres. sg. 3. frisalethar, Sg. 2078 3; cf. Ml. pres.sg.1.nqrithaZim,non 1298 2b expecto, M1. 49d 3. 4. pl. 1. frissailemmarni, g. praestu- lamur, M1. 63c 7. pret. sg. 1. frisrailadur, expectavi, M1. 86‘3 8. fut. sg. 1. frhailefwsa, g. praeetulabar (1.-bor), M1. 388 10. *airliur, studeo ; con-airliur, condo. pres* f airletirar, Wb. 17b 22. sg. 3. conairlethar, g. consultat, M1. 12P 2. pl. 1. lase oonairlemmarni, g. durn consdimus, M1. 18C 1. subj. sg. 3. manicdinaidither, Wb. 5b 38. arndip maith n-airlethar, Wb. 28b 32. pret. Bg. 3. lase wairleat&, g. consulendo, M1. 1250 1. cluiniur, ro-cluininr, audio ; suh junctive cl6ar.l niscluinsthar, non ea audit, sg. 3. 1 MI. 21b 2. roclzcinethr, Wb. 12~22. pl. 3. racluinstar (1.-ethar),g. audit, Ml. 1148 13.

1 This difference of stem for indicative and subjunctive has been pointed out by Thurneysen, KZ. xxxi. 80, for the language of the Glosses. The deponent cZm- forms which I have noted in the early profane literature are indicatives. CZo- is found as a subjunctive, LU. 114b 23, coclothar tnrluag ‘that the host may hear,’ and in the new analogical forms, which have arisen from the 1 sg. cldor, cMr, dacldra n’ ‘if the king hear,’ LL. 46a 7, m’cocldrat, LL. 96. 18. Where the deponent idexion has disappeared clan- also appears in the snb- junctive, as dianathcluine ‘if he hear,’ LL. 27Ib 33. Cldm is best taken as =*chdr, *c&dr, the subjunctive of an a aor. ; for the ending cf. msisur. 450 THE DEPOKEST VERB IN IRISH-J. STRACHAN.

DEPONENTFORMS. ACTIVEFORMS. rodchluinethar, qui id audit, Wb. 12c 46, cf. 27b 27, 2ga 18. rodchluinethar, g. auditori, MI. 129c 19. pl. 1. an rochluinsmmar, MI. 112b 13. 3. anrochliiinetar, podaudiunt, Wb. llb 6. subj. sg. 1. niconchloor, non audiverim, Wb. 23b 41. acht roeloor, dummodo audiam, Wb. 23a 2. resiu rocloammar,' antequam audiamus,Ml.l12b 12 (bis). ipv. sg. 2. cluinte, M1. 1368 10. fut. ag. 3. rotchechladar, qui te audiet, Wb. 28d 16. rondcechladar, MI. 53b 27. *comalour, impleo, fulfil (denominative from comldn, perfectus). pres. ind. nodchomalnadar, qui implet, comrnallaid, M1. 106s 2, sg. 3. 1 Wb. 15b 14. cf. 94b 1. hore comalnas, Wb. 7b 15. pl. 3. comalnatar, qui implent, Wb. comallait, MI. 146s 1. 20d 2. nl comalnat, Wb. 20~22. forswahicomalwatar, Wb. 20a 1 horanadconaalnat,Wb.3lb 31. intain naeomallatar, Ml. 94b 6. chomallaite, M1. 114b 7. subj. sg. 3. maniscomalnadar, Wb. 2ga manicomalla, Nl. 9ib 10. 16. act rachomalnathar, Wb. 31b 11. arinchomalnathar, Wb. 3 lc 14.

* As a pass. rocloatar, M1. 708 2. Stokes translates cochmrchloatar, g. ut gratiae sgantur ro nobis, Wb. 14c 6, by 'that they, may be heard(?)' but the word is probatly to be put with coerchl6i, g. feritatis eorum et truculentire crimen exagitat, Ml. 33d 16, conzirchloiter, g. agnntur, Wb. 48 14, cf. M1. 18d 19, Sg. 17' 7. In our passage it seems to be a mechanical tranelation of ayantur. THE DEPONENT VERB IN IBISH-J. STRACHAN. 451

DEPONENTFORMS. Acrrv~Foam. corrslple'ir roacomallathur, MI. 12gb 2.' ndd chomalnathar, who shall not fulfil, Wb. 21c 14. pl. 1. cocomahammar, Wb. 3lC11. diacomalnammar, Wb. 15a 27. manoacomalnnamar,Cod. Cam. 38b. diand comallamar, Ml. 460 20. 3. manicomobatw, Wb. 2c 10. machomalnit, Wb. 28~7. ipv. sg. 2. noacomalnithe, Wb. 30a 1. pret. sg. 1. arrocomallua, MI. 74a 5. 3. rachomalnastar, Wb. 24s 37. roecomaZn[aatw], Wb. 21b 9. rondchomallastnr, M1. 122a 7. PI. 3. nirwcomalarLatar], MI. 1058 6. rochomalnieset,Wb. 2c 14. fut. pl. 3. acomallaibte, M1. 121b 8.

cniriar, pono. pree. sg. 1. doczciriur, g. ascisco, Sg. 16b 8. 3. adwirethar, redit, Sg. 73b 2. docuirethar, g. asciscit, Sg. 61a 7. doczcirethar cktna penan sin persana aili chucae, i.e. asciscit, Sg. 1918 2. hnderbeirt Biuth dochuirethar, g. usu invitante, M1.29b 1. amal dzcndchzcirethar,2MI. 35d 22. anduduiredar, Bcr. 32C 10. andi doczcirethur, Bv. 4d 1.

1 For the mood cf. coru[p]Uirdungnd ' so that it be industriously that he does,' M1.129b 1. 2 niradi . . . . amal dmdchirethar ima 6a&, lit. 'he does not speak as happens to come into his mouth.' For this intransitive use cf. Bcr. 32c 10, anduchuiredar bissext ' when the bissext falls,' and the very fragmentary glms, Bv. 4* 1. This usage is also found in the profane literature, e.g. LU. 92. 30, LL. 248' 24, 275b 7, 186" 49, 1878 66. 452 THE DEPONENT VERB IN IRISH-J. STRACHAN.

DEPONENTFORMS. Ammi FOBMS. andonaithchuiredar, redit iterum, g., .u. redeunte in vocalem, Sg. 18b 6. amal duadchuridur, redit, M1. 34a 7. doetarcuirethar, g. interpellat, Wb. 4b 19. neich fritcurethar clrsill, qui colit, Ml. 41d 16. pl. 3. hicuiretar salt, in quo ponunt saltum, Bcr. 32b 11. adcuireddar, redeunt, Sg. 202b 3. n%erchuiretar,’ Sg. 6b 17. immechuretw, transfeerunt, Wb. 58 5. subj.*sg. 1. 1. C0fri8tUEW,Sl\dl.11 80 3. 2. nliscoirther, Wb. 108 21. cuiri huait, g. tolle, Bcr. 328 11. 3. mnach corathar, MI. 68b 9. condichet andruim, ut ter- gum vertant, Ml. 448 16. ipv. sg. 2. mirth, MI. 56C 5. frecuirthe cheill, g. recole, Bcr. 33b 2. pret. sg. 3. dorrochu~restar,g. arcessivit, sg. 1848 2. dorochuirestar, g. exciverat, MI. 16O 6.

1 “on annnllino,’ Ascoli. It Beems to stand for *ma-ro-churetar. 2 dichret seems to stand for *dZ-cerGnto, for *dZ-corcinto, or *Q-core@nto, might have been expected to give Wichrat. This subj. stem cerci- would be like be&, mcnii- (KZ. xxxi. 78), only here the indicative has a causative formation, “coreici. In -cor, -corathar, to-chord, MI. 441 19, cor- may have been substituted f& cer- under the influence of the indicative forms. In Ml. 72d 1 dunathfoichret ‘ they will return,’ stands for do-n-ath-fo-cicerrat, from do-ath- fo-eherdaim. Cf. lase donaithfocherr .i. dew, M1. 34d 8, where it seems to mean ‘when he (God) will bring back,’ unless we suppose that the glossator has misunderstood his text. Cerd- appears intransitive in W. cerdded ambulare. Cer-d- may be looked upon as an extension of cer- of crciriur. In M1. 78c 8 cerdaim seems to take the place of the usual cuiriur, frisfoichiurr mill g. incolam. 3 Read cofristaeor that I may cast back upon them,’ cf. friatacuirther, sg. 21b 4. THE DEPONENT VERB IN IRISH-J. STRACHAN. 456

DEPONENTFORXS. ACTITEFVI~XS. drothuchurestar, g. non exciverat, M1. 18d 6. dorochuiristar, g. adsciverit, MI. 25C 13. dicdtoeJarastar,l M1. 39. 3. ducorastar dia hilb rnoraraige . . .foreinnf Bad,Nl. 5 2, pl. 1. doroohuireemmar, g. ascivi- mus, Sg. 6b 18. fut. sg. 1. hczcirifar, g. citabo, M1.38 1. iiimeccim, dispicio? pres. id. demecimm, g. detero, sg. 1. 1 39b 1. 2. hmmeiccither, g. inhonoras, Wb. ld 13. pl. 3. amal dundatmecetarsu, ut te dispiciunt, MI. 106c 11. subj. sg. 3. arndimicedar, g. neinhonoret, M1. 122d 2. wudimicedar, MI. 1298 14. pret. sg. 3. nf rodimicestar, Ml. 119. 10. *emu, fut. of ed- ‘eat.’ The presential tenses are supplied by ithim. Bg. 3. ciaestir, etsi edit, Wb. 6b 24. ciniestar, etsi non edit, Wb. 6b 23. remiais~idider,~g. praefatur, MI. 640 9. fetar, scio.( sg. 1. roftar, Wb. 9b 17, etc.

1 The MS. has according to Mr. Stokes dia corastar. Ascoli’s correction is required by the sense, ‘ Saul came upon him.’ 2 The remaining forma in the Glosses are afidm,.mecitis,dum me dispiciebant, M1. 34~6, diinicctther ‘he is reprobated,’ Wb. Sb 16. The noun is dimiccem ‘ contempt.’ The first part of the word is di-, the rest is obscure to me. J The forms of the verb asilzdiest, nmo, are regularly active. The deponent form ma here, perhaps, be ascribed to the influence of the Latin word. Or did the glossator think that he was translating a passive ? 4 Fetar, an s aorist from 1/cvid (cf. Thurueysen, KZ. xxvii. 174), supplies the present indicative ; the pres. subjunctive comes from *j;nn- =Skr. vindimi Phil Trans. 1891-2-3. 30 454 THE DEPONENT VERB IN IRISH--3. STKACHAN.

DEPONEKTFORMS. Acnvx Forms. cer&?jitarsa,Nl. 55d 21. 3. rojtir, Wb. 7c 15, etc.; cf. Xl. 27d 7. subj. sg. 3. toneich rudjnnadar, Ml. 46c 24. pl. 3. amal nad$nnata/r, Ml. 99b 10. 8. fut. sg. 1. rofessursa, Wb. g8 21. 3. n$astw, Wb. 12a 18. ruJiastm, Ml. 11 Ic 13. rofestw, Wb. 12d 27. pl. 3. rofessatw, Ml. 69b 1. 8. subj.sg.2.3feser, Wb. 298 22. gfeiser, Sg. 209b 30. 3. aonfestm, Wb. 12c 38, ete. pl. 3.’ cofmatar, Wb. 26d 33. yfociallur. pres. ind. fuciallathar, g. procurat, Ml. Sg. 3. f 114b3.2 arohhladar, Sg. 209b 26. Cf. ZE. 981 note, where the verb is explained as a denominative from focuL2

f~llunr,~g. VO~O. Sg. 146b 11. pret. pl. 3. orufoluaesat, g. convohsse, M1. 670 16. Yolnnr, rego. pres. ind. folZalalthersu, g. regis, MI. sg. 2. f 82d 5. sg. 3. fallrzaither, g. regi, i.e. qui regit, Nl. 90% 9. ipv. sg. 2. follaids, g. rege, M1. 46b 18. Fret. sg. 3. ~rwfollnadar,Wb. 13b 29.

‘I find,’ which also upp plies the secondary present, rodJinnad, Sg. 209b 25, and the imperative &nad, Wb. 298 17 ; in Mid. Ir. the 2 sg. ipv. is regularly finnta. In the asske ro$nnatar, Wb. 2ga 28. ForJilani- as subj. stem cf. also eo@anamdr, Ll? 112b 32. As fetar is regularly deponent both in Old and in Middle Irish, I have noted only such forms aa seemed of interest. 1 Cf. foehiallastar .i. rotinbiz ‘he collected,’ LU. 132 1. 16. 2 Cf. ~ifwlamw,LL. 62. 14. 3 The pres. sec. fuluinn, g. avolare is found Ml. 140b 8. THE DEPONENT VERB IN IRISH-J. STRACHAN. 455

DEPONENTFORMS. ACTIVEFORMS. fut. sg. 2. folnaibe, g. reges, 311. 188 4. PI. 2. fokebthe, regnabitis,Wb. 25b 25. *gainiurll nascor, Jcaithgainiur, renaacor. nicongainidar, Acr. 61. sg. 3. innahi hua+gainedar,Sg.139b 4. idan . .. gainsthar, M1.448 11. pl. 1. adgainemmar&, M1. 66b 6. 3. huangainetar, Sg. 39%26. perf. sg. 1. cwogenarea, M1. 440 11. 3.2 Togenair, MI. 24a4; of. 25b5, Sg. 318 6, Wb. lla17. ad-gladur, alloquor, imm-ad-gkddur. pres. ind. 1 adgkddur, g. appello, Sg. sg. 1. 146b 9. 3. adgladathar, g. adloquitur, MI. 748 8; cf. M1. 838 9, 115d 1, 1208 8, 143 14. adglcidathar(rel.),Sg.211bll. MEngl[ad]adar, g. dum con- venit, Nl. 53c 17. PI. 3. imrnusaccaldat, g. motuo se adlocuntur, Ml. 131c 19. subj. sg. 3. adgladadar, g. inducitur con- venire, M1. 1158 6. pl. 1. act immanrirladmar,3ut mutuo no8 adloquamur, Wb. 29d 10. *labm, loquor. intan nokabrith, Sg. 1598 2. sg. 2.

1 Correa ondig to an indicative gm- we should expect in the subjunctive gen- (cf. $hurnepen, KZ. xxxi. 78). I have noted no instances of the subjunctive, but the indicative in Mid. Irish is genithir, or with change to the active inflexion genid. Now gm- cannot be equated with gan- of the Glosses, and the probability is that gen- here comes from the subjunctive stem. * A peculiar 3 pl. rogkartar is found, Wb. 4c 12, cf. ZE. 1091, where rogdnatar has become roghndrtar nnder the influence of the sg. rogdnar. 3 =imm-un-ad-ro-gWtnur. 456 THE DEPONENT VERB IN IRISH-J. STRACHAN.

DEPONENTFORMS. ACTIVE FORMS. 3. ni Zabrathar, Sg. 199 6. labraid, Ml. 1 15a 2 (bis). ileti labrathap., Wb.12c 19; cf. 24C 29, 7d 2. pl. 1. labramarni, Ml. 31b 23. lase labraimms, M1.3 Ib 15. 3. nilabratar, Wb. 12b 22. frislakatar, g. oblocuntur, Ml. 55a 10. inta7a labratar, Sg. 162&3. subj. sg. 1. labrar, Wb. 120 36. 3.‘ conicsm nadlabrathar, Sg. 199b 5. pl. 1. labraimme,2 Ml. 31b 16. 3. cenwlabrafar, Wb. 12a 28. nabad immalle Zabritir,a Wb. 13a 5. pret. sg. 3. rodlahrastar, Ml. 126C 10. rolabrastar, ib. fut. pl. 1. ni Zabrafammar, Wb. 12~4.

laimiur, rolaimiur,4 audeo.

sg. pres.1. rzclaimur, Wb. 17c 21. rolaumur, g. audere,Wb. 1788. rolomur, g. mihi audenti, MI. 21b 5.

1 In a couple of instances it is not clear to me whether labrathar is an indicative or a subjunctive. Wb. 13. 4 mabeith nech and tra labrathar might be interpreted ‘ if there be anyone who speaks,’ cf. masw tho1 atonzaig if it be desire that drives me,’ Wb. 10d 26, or ‘who shall speak’; cf. naad-lleteg dogneid, Vb. lld 16. In M1. 35d 22 eondcb sainemail taannf bbrathar we seem to have a subjunctive ‘ so that wbat he shall say (&r hv dm)may be excellent.’ In Sg. 19gb 7 is inderb in prima nodlabrccthar, 1. tertia, ‘it ‘is uncertain whether it is the first person who speaks it or the third,’ we seem rather to have an indicative. ’I The form of the sentence is in favour of taking the verb as a subjunctive- eorbzk dureir nachaili labraimme ‘ that it should be at the will of any other that we should speak.’ 3 For the mood cf. bad hed dogneithsi, Wb. 12d 7, also Wb. 13” 6, 29. 4 With this verb Ascoli puts foldmur, suscipio, tento, confolmaissiur. MI. 50d 8, foiizkimastar (MS. fomkrimas COR. Stokes), Wb. 17d 4, arfolmastar (for arfolmas, Stokes), Wb. 218 2. In Sg. 147b 4 is ecentech ingnim arafolmathar, where dscoli takes the verb as a passive, it seems rather to be active, ‘ the action which it undertakes is indefinite.’ But the inflexion of the verbs is different, and we have perhaps rather in *foldmwr a derivative from lam ‘hand,’ drrxapb. THE DEPONENT VERB IN IRISH-J. STRACHAN. 457

DEPONENTFORMS. ACTIVEFOBMS. 3. rolaimethar,' g. audet, Wb. 5' 5' 15. pl. 1. rolaimemmar, g. audemus, Wb. 15c 19. niconlaimsmmarni, Wb. 17b 8. subj. sg. 1. rolZdmw, g. ausim, Sg. 171b 1. pl. 3. na laimetar, g. non audeant, M1. 60d 6. midiur, iudico.2 midiursa, Wb. 98 2. sg. 1. 1 2. fwsammitter, de quo iudicas, Wb. 6b 22 ; cf. 10 10. 3. naithir, g. diiudicat, MI. soc 8. nimidedar, non cogitat, Sg. 63" 14. ish cotantmidethar, Ml. 17 b 2 (= cot-dam-). conmidethar, g. principatur,s M1. 121c 18. duntmihthu~,g. adpendit,Ml. 82' 3. ni irmadadar, Wb. 288 21.' pl. 3. innahi immeruimdetw,6 qui peccant, Ml. 46b 23.

AscoIi correcta na&adme&, Wb. 170 19, to ralaimedar. 2 The Idg. root med- appear8 in Irish as in other Idg. languages (e.g. Gr. pWopt, pc6e'ou, p@opai) as med-, mid- =Idg. mU-. Med- is found in the 8 forms, mid- in the perfect forms. When the accent shifts back to a preceding syllable mZd- becomes mod- @nhdeed, but fbgnam), thus ramtdar but cona'mmadar, rumidar but dodmadir. In Tur. I7 immarwmedair is peculiar; if it be not a blunder it might be explained as an analogical transformation of *dmm&air after the mm-forms. a From *con-midiw comes apparently lzad coimmestar, non potuerit, M1. 1278 19 ; cf. coirnadiu, dominus. In W. meddu, posse, we have the grade med-. 4 The meaning of the gloss is not clear. Stokes translates doubtfully 'he does not understand.' The compound armidiur is found LL. 2908 14 dranevmasta, where it seems to mean 'if it were tried' ; cf. conartnadatar degcomairli, M1. 646 11. 5 From iname-ro-midiur. On this compound see Windisch 1%. Forsch. 111. 73. 458 THE UEPOKEKT VERB IN IKISH-J. STRACHAN.

DEPONENTFORMS. ACTIVEFORMS. ni irnaadatar, Wb. 5b 2.' subj. sg. 3. comidithir, ut iudicet, MI. 35c 30. pl. 3 (as na aimdetm,3 g. non conantur ipv.).2 (1. conentur), hil. 56' 23. perf. sg. 1. rmidm, eum iudicavi, Wb. 9b 5. conammadarsa, adjndieavi, Wb. 26b 21. rumidair, g. duxit,Ml. 72b 21. imrumadir, peccavit, W b. 13b 31. immarulloediar(leg.-aik),Tnr. 17. itheside dorumadhi, g. que fuerat emensus, M1.16c 1 1. pl. 3.4 eonnarmadatm &gconaairli, Bill. 54d 17. anduruimdotm, M1. 87" 4. s. fut. sg. 1. dummesaur~,g. metibor, MI. 788 11. 3. miastar, pi iudicabit, Wb. Id 9. nmnaiastar, g. examinans, ML 56s 10. intan imnaeromastw, M1.5 18 18. pl. 1. messimirni, Wb. 9e 10. nosmessammar, Wb. 9C 10. pl. 3. kroimset, g. delinquent, M1. 54a 23. irnruimset, g. peccabunt, M1. 548 27. s. subj.sg.1. inmesw, g. si nanciam, Sg. 1798 1.

1 Stokes translates here by 'attain'; cf. condrrmudair (hgai) '80 that the spear hit,' LU. i3b 20 =*&ro-midair. 2 The ipv. is otherwise expressed, Wb. nacAibmi&d, g. nemo vos iudieet, Wb. 27' 24. 3 From ad-midiw; cf. ammzcs 'an attempt'=*admesszrs; cf. also dammi- dethnv 'he tries,' LU. lllb 45=*to-ad-mid-. 4 So Ascoli for conaconnamnadatar. THE DEPONENT VERB Ih' IRISH-J. STRACHAN. 459

DEPONBNTFORMS. ACTIVEFOBME. amal nummesw, g.iudicaturus, M1. 94b 8. 2. na imroimser, Wb. 200 4. 3.' w8tarl g. iudicandi, MI. 127d 12. lase dumestar, MI. 68d I. arna imromastar, Wb. 1 la 16. adrnachduq2 g. miror, Sg. 68 4. -moiniUr, subj. -menar. pres' "" } domoiniur, puto, Wb. 148 10. sg.- 1. domzcinwsa, Sg. 20gb 10. 2. domointerso, g. existimas, Wb. lc 13. 3. doaithmiwdw, g. commonen- tern, Ml. 136c 11. taidmenadar, comminiscitur, mb. gC 30. taidmincdarsom, Sg. 2Zb 10. ni taithminedar, Sg. 13b 4. dunaithmenadar (rel.), M1.52. foraithminedar, memorat, M1. 17b 23. fodaraithmine[dar], qui eius meminit, MI. 25c 5. is& 808 foraithminedar, MI. 55c 1. arasmuinethw fiid,' honestat earn, Y1. 368 18. ardamunethar feid, qui ea veneratur, M1. 124C 16. farmuinethar, g. fascinat, Phil. 10.

1 Here I should be inclined to put =nad coimmestav, g. nequierit, Ml. 127' 19, which Ascoli derives from coniecim. But coniccim, except in the perfect, seems in the Glosses to be regularly active; cf. Ascoli, Lex., C. That coimmestar might go in meaning with md- is shown by commus, power= *COIIIAnedSZCB. 2 The full gloss runs admachdzcrsa .i. isnaachdad Umm. The latter is the idiomatic Irish translation ; the former is evidently a word coined for the occasion to express mirw. 460 THE DEPONENT VERB IN IRISH-J. STRACHAN.

DXPONENTFORXS. ACTIVEFORMS. pl. 3. dodmainetar, qui putant hoc, Sg. 58 6. dummunetar~om,Xl. 113d 1 1. damunetarsom, Ml. 35b 10. indi domuinetar, MI. 39a 26. dommuinetap; M1. 4gb 7. for&6moinetlor, invident vobis, Wb. 19d 27. subj.'sg. 2. fom.ntar,z Wb. 28C 9, 30C 2. 3. dummenathm, g. suspicari, Ml. 4g8 15. na tomnathm (as ipv.), Wb. 178 13. arnatmn(a)thw, M1. 30C 5. arnackdermanadm, ne obli- viscatur, M1. 3Zd 5. pl. 1. dumenamrnar, g. opinari, M1. 7Sb 24. andurnenmami, 1\11. 15d 5. arnatomnammmni, Nl. 15a 5. condermanammarni, ut obli- viseamur, Ml. 21e 3. pl. 3. in& fw&%menatm-, ,111.34.23. intomnatar, g. putent, M1.18'5. cofummenatar, g. ad cavendum, Ml. 438 16. aformmatar, g. invidentes, M1. 17b 16. ipv. sg. 2. nachamdwmainte, M1. 326 5. perf. sg. 1. dombama, putavi, Wb. 3= 26, 27. durumenar,N1.49b 13,130d4.

In Wb. 310 10 marmentar fdid is found glossing ut revereatnr, but with the form of the pwive; cf. armmentar fe'id, g. suscipi, MI. 61. 11. Has the glossator misunderstood his text, or does armentar stand to arwmathar 8s dummmar to dmenammar? Considering how carefully the distinction between deponent and passive fom (e.g. foraithminedar 7 fwaithmeteter, MI. 17b 23), the former alternative seems the more probable. a -mentar here is interesting as being the regular phonetic development of =*mendter. In other cases where -ter might have to give -tar, it has been preserved by the influence of other verbs, where it re arly remained because of the slender nature of the original vowel in the preceeg syllable. THE DEPONENT VERB IN IRISH-J. STRACHAN. 461

DEPONENTFosnrs. ACTIVEFoms. ni ~tormenarsa,Y1. 428 10. tommCnsa dial putavi me deum (1.-mknarsa), Nl. 130d 5. 3. niruforaithmenair, Ml. 24c 8. cEmzlmenairsorn, putavit, M1. 61d 2. dorumenareom,’ MI. 32d 10. d~ar~rnuinestar,~Wb. 4c 38. pl. 1. ni tormdnmmni, M1. 115b 1. ni dermenmarni, g. obliti non sumus, M1. 648 3. 3. dorumenattar, M1. 35b 18. foruraithminset, Ml. 1358 1. nf tormenatar, Ml. 900 5. wrumuimet fc;]id, Ml. 908 1. nadtormnatar, M1. 95b 3. fUt. Eg. 1. aratmuinfwsa fdid, nam te venerabor, Ml. 63* 3. 2. ciafut dunhroimne- fese, puouaque me oblivisceris, Ml. 32d 5. pl. 3. arrammuinfetw feid1Ml.6la16. duroimnibetar, obliviscentur, Ml. 778 12. molnr, laudo. pres. ind. hore nondobmolorsa, quia VOB sg. 1. f Iaudo, Wb. 140 18. 3. nodmoladw fesin, g. qui s6 ipsum commendat, Wb. Hb 21.

1 Ascoli translates dorummar am dia diadermut b ‘oblii, Iddio per la smemoraggine [tmcuranza] ma,’ but in the text Dad does not say that he has forgotten God, but asks how long God will forget him. The gloss would seem then rather to mean ‘he thought that God had forgotten him,’ which in Mod. Ir. might be expressed by the same idiom, shaod se‘ dia do bheith dd dhearmhad. I have placed this form here with Bome hesitation. Stokes translates the gloss innahi diarecmuinestar aom trocairi by ‘those unto whom he taught mercy,’ taking the verb from mliinim ‘teach,’ but the context seems to require ‘those to whom he intended mercy ’ ; rumztineatar would be a new formation, a dep. E. pret. 462 THE DEPONENT VERB IN IRISH-J. STRACHAN.

DEPONENTFORMB. ACTIVEFORMS. pl. 3. cid ariptdmoZatar, Ml. 1456 4. ni molat, Ml. 130b 6. subj. sg. 3. condidmoladar, Wb. 16d 1. moladar, g. commendet, Bcr. 41d 1. pl. 3. co molait, Ml. 51c 10. pret. sg. 3. rommolastar, Ml. 126b 16. fut. sg. 1. nomoZfar, Wb. 98 22. atmuilniur.l g. iterum dico, Wb. 18C 12. *condirgnr (= om-di-reg-), dirigo, corrigo, "tremedirgur, transfero. pres. ind. condirgedar, MI. 343 23. sg. 3. ] aoondirgedar, g. diregene, Ml. 468 12. trimedirgedar, g. transfert, MI.. 548 22. diatremdirgedar, Sg. 1908 5. pret. sg. 3. arrudirgestar, Wb. 4c 13. fut. ag. 3. oondirgebadar,g. correcturum, MI. 130C 15. SamlUr, imitor (samail, similis) ahd-8amhfr,2 aemulor, di- samlwr, dissimulo, di-ind-8amhr, id. pres' indm is friss pta eamzur, Wb. 3c 6. sg. 1. 1 3. disamlathar, g. dissimulat, Ml. 21b 2. h~dathar,Wb. ga 15. insamlathar, sg. 30a 14. ni de intamladar, MI. 2V 12. pl. 1. inintsamlammarni, g. bnaemul- amur, Wb. llb 16. coni intamlammar, g. neque emulamur, Ml. 97c 1.

1 This verb seems as yet to have been found only in this passage. It could be analysed into 6th-d-muilniur ' 1 say it again.' Does -muZniwr come from *mohendr, cognate with mokw? For the two meanings cf. Gr. aivL. 2 Before the accent the B remained maspirated and itad-a- became inr-. After the accent a became ah, which caused the change of d to t. THE DEPONENT VERB IN IRISH-J. STRACHAN. 463

DEPONENTFORMS. ACTIVEFolrrs. subj. sg. 1. ciinsamlar, g. si imiter, Sg. lb 1. 2. insamailter, g. noli cmulari, ni iBtarnlae, Ml. 56c 13. M1. 56b 38. pret. pl. 3. arinrwamlasatar, Sg. 112b 4. fat. sg. 3. ni deintamlafa, g. non disimulabit, hI1. 6gd8. pl. 1. nonsamlafammar, Wb. 17b 12. *sechur, sequor. pres. ind. andocoisgedar, g. sequente, sg. 3. f Sg. 1~5~4. dudchoisgedw, (ROX) quae se- quitur, M1. 62C 6b. otochsgedar, g. consequens, ML. 103d 25. pl. 3. sechitir, Wb. 31b 20.‘ ni sechetar, Ml. 19b 11. dudchoisgetar, M1. 17c 12. subj. sg. 3. airesechethar, Cod. Cam. 37c. madocoisgedw, Sg. 16b 2. pl. 1. arasechemmw, wb. 2d 4. pret. sg. 2. or~thochaisgesswsu, g. con- secutus es, M1. 430 9. 3. roiechestar, Sg. 2OBb 15. duruchisgestrur, g. persecutus, Ml. 64c 8; cf. 98b 7,99b 11. fut. sg. 3. fomthochaisgebadarsa, g. subse- quetur (mihi), Y1. 456 5. dofuislim, labor. pres. ind. dofuislim, g. labo, Sg. sg. 1. 1 146b 1. subj. sg. 2. hotuislider, g. quo labeberis, M1. 2d 6. 3. dof&sledar, g. elabi,Ml. 300 10.

sechitir goa ‘they follow lies.’ For the construction cf. Nl. 89c 5, aeichfedsom du$rinnieiu ‘he would follow thy truth,’ LL. 346a 50 s.&hed$ri*mi nosseichfe ‘ let him follow truth that will follow him.’ The strange form numseichethve, sequatur me, Cod. Cam. 37c, should probably be corrected to numseichethse, as ZE. suggests. 464 THE DEPONENT VERB IN IRISH-J. STRACHAN.

DEPONENTFORMS. ACTIVEFoaaas. dufuisledor, g. relabatur, Bcr. 34c 1. fut. sg. 3. nicontuislifsa, M1.27b 18. -sissiur,' i'mapar. pree. ind. anarmbsiursa, g. innitens, MI. sg. 1. 1 41C2. duair(i]siur, innitor, 1161. 798 9. huare doonairi88iul; MI. 790 10. 2. anduneriasi&rsu, g. te ad- etante, Ml. 38c 25. 3. lase araalsedar, g. cum inni- titur, Sg. 213b 2. hi8 in&forsanairissdar inynim, Sg. 139b 1. oterissedar,constat, M1.14504. amal nadtairissidar, ut non consistit ( subj. ? ), Ml. 104b 4. fosisedm, qui profitetur, Acr. 69. pl. sg. 3. arasissetar, g. innituntur, Ml. 39b 11. fristairissetar, g. obstantes, M1. 47c 3. fosissetar, confitentur, MI. 1320 1 ; cf. Tur. 43. subj. sg. 2. fosisidarsu, Ml. 66C 1. 3. lase arasissedar, g. cum fuerit innisa, M1. 51b 13. mani airissedar, 511. 131b 4. co arsissethar (ms. -etar), g. ut nitdur, M1. 120a 19. pl. 3. arasissetar, g. niti, M1. 77b 3. conitairissetar, ut non con- sistant, Ml. 104b 4.

I When the accent falls on the verb itself the initial I of sissiur is preserved , if the accent be thrown back on a preceding particle, the s vanishes if the immediately preceding originally ended in a vowel. A further instance if;;:% ting this rule is noaraiaainn, M1. 44a 26. THE DEPONENT VERB IN IHISH-J. STRACHAN. 465

DEPVNENTFORMS. ACTIVXFoam. fri8tairissetar, g. qui obsistere non verentur, MI. 39b 13. fosisetar, g. fateri possunt, 8g. 140' 2. pret. sg. 1. awoissiursa, g. nisus sum, Ml. 88' 9. 3. arroiseetar, g. innisurn, M1. 18d 16. afwoiseestm-, g. confessus, M1. 46d 15. nitawaetair, Acr. 72. fut. sg. 1. fosisefar,confitebor, M1.58c17. -tluohnr. pres. ind. atluchur, g. gratias ago, Wb. kthluchimse, g. efflagito, sg. 1. 1 3b 19. Ml. 71c 20. duthluchur, efflagito, Pcr. 578 5. ma1 dunthlaichiursa, M1.44c 20. ma1 duntluchur, XI. 74a 3. 2. cia aratodlai[g]ther, g. quare postulss, Y1. 3Za 5. 3. atluchedar buidi, Ml. 128C 3. dutluichethur, g. exigit, 1\61. 36.28. duthluchdar, Ml. 38d 1. pl. 1. oldaas duntlucham, Wb. 21d 9. 3. duthlucktar, exigunt, MI. 85d 5. subj. sg. 1. coduthluchar, g. ut efflagitem, Ml. 49d 10. cia$% kdlaigersa, g. quam iusta postulem, Nl. 3Sc 22. pret. sg. 3. rotodlaigestar, M1. 124d 9.' fut. pl. 1. atluchfam, Wb. 178 1.

1 duthluich in the same gloss should probably be emended to duthZuiehe6 huare ambert cia duthluich[ed] naditatais ' because he said though he did ask they would not get.' 466 THE DEPONENT VERB IN IRISH-J. STRACHAN.

DEPONENTFORMS. ACTIVBFORYII. dufo-thraccar,' opto, volo. perf. sg. 1. dianduthraccama, quibus opto, Wb. 14b 5. 3. ciadudfutharcair, Ml. 52. pl. 3. dufuthractar, g. volunt, Wb. 26b 1 ; cf. Ml. 498 17. doauthractar, Wb. 26b 1. 8. fut. sg. 2. nf di~thrais,Carm.Ml.17. pl. 3. huare dunfutharast, MI. 548 28. s. subj. doduthris, 20b 9. sg. 1. m. doficthriaae, Wb. 328 9. pl. 3. ci dutairsetarsom, MI. 56c 7.

In the following verbs deponential forms are found in only one part of the verbal system.

ad-cin ' I see,' and other compounds of *-ciim ' see' (= %esi6 dkes). Here the subjunctive mood is deponent 2 pres. subj. ma fri~accar,~si sperem, MI. sg. 1. t 49d 3. 2.' adaiicider, g. (cum) respicies, MI. 438 19. sechidh dsnecaitherw, g. qua- quam mrsus respexeris, MI. 73c 11. diandadercaithersu, g. si eos aspicias, M1. 109 10. 3. odidaccadar, Wb. 16d 6. coniaccadar, g. quin videat, M1. 538 6.

1 A perfect form with a present or perfect meaning. a Except, of course, in the 2 pl., where the Irish verb has no special deponent forms, mu adced ' if ye shall see,' Wb. llb 22. 3 The orthotonic form is found in a poem from the St. Paul Codex (Windisch. Ir. Text, I. 316), rop ith 7 mlicht dear 'may it be corn and milk that I shall see,' I. 1. 4, similarly 1. 6. 1 In an imperative sense an active form is found, &me Zut, Wh. 100 20, d62iCCSSit4, Wb. 31b 19, and often in the profane literature. THE DEPONENT VERB IN IRISH-J. STRACHAN. 467

wnaccathar, Ml. 68b 9. pl. 1. manidecammar, 8g. 26b 9. 3. frianaccatar, g. sperandi,' Ml. 1248 3. coniccim possum.) The perfect is deponent. sg. 1. cotaneccaraa, id potui, Wb. 143 40. 3. conanuczcir, MI. 1197. niconclwimnzcczlir, Wb. 19c 10; cf. MI. 116c 5, 97b 4, 97d 4, 5, 10, etc. forcomnucuC,1 Wb. 22b 8 ; cf. Sg. 1488 6, MI. 131b 14, etc. tecomnzccuir, Wb. 108 4 ; cf. Cod. Cam. 38e. fodamaim, patior, perfect dep. so from ad-damaim, concedo. sg. 1, foarodamarsa, M1. 3g8 13 ; cf. 13, 1328 12, 95d 13, 14, Wb. lgd 10. drodamar, Acr. 99. 2. adrodamarau, Acr. 29 (MS. aa'romarsu ) . dhinndrdamarsu, Acr. 46. 3. forodamuir, Ml. 548 25; cf. 54b 28. eiassair, sedit, daed-; cf. Zimmer, KZ. xxx. 151, Thurneysen, KZ. xxxi. 97. im[m]mimaair, g. obsidit, M1. 43b 1. Isolated deponent forms are coinedar, plorat, MI. 74b 1 (actiye forms are common, e.g. M1. 73b 12, 86d l), immetha- crathw ' covers,' MI. 6.P 1 ; cf. p. 68, note.

1 There is here a double gloss, cofrisaceat and frisnaccatar. The former I take for an indicative ' so that they expect,' the latter for a subjunctive ' that they may expect'; cf. MI. 4~5~17, where cid frisnaecat glosses sperandi. For other examples of the present indicative, see MI. 53'3 23, 62d 7, 69b 5, 1348 6, 144* 2 ; Pcr. 64b 1, 678 7. * mdartor, Wb. 7' 14, is obscure. 468 THE DEPONEKT VRPH IN IRISH-J. STRACHAN.

In the following forms isolated deponential forms are found in the 3 sg. of the preterite : rodligestar, M1. 368 29, &ligim 'I owe,' like derivative verbs in -igim. arruneastar,' M1. 50b 8 amsithim, sustineo, expecto. adroneestar, Wb. 4c 35 ] arruneilhstar, MI. 638 15. In sg. 1 forderisiur, g. lustravi, MI. 133b 8; cf. fwfhet, g. illustrat, MI. 78b 8.

The second class of deponents consists of denominative verbs in -agiur, -igiur. The origin of the formation is uncertain; cf. Brugmann, QrunlEriss 11. S 780. In Irish the fdrmation has become closely associated with adjectives in -ach and -ech, -ag- as a rule appearing by -mh, -ig- by -ech. But such verbs are also found with no corresponding adjectives by their side. From cases like "Zethnaigur, "bresminigwr, where the n belongs to the stem from which the verb is derived, -nag, naig-, -nig- has spread as a suffix, e.g. bruthnaigim, "freptanaigur, "mrechtnigw, and bertnaigim (cf. Windisch Warterbuch) by bertaigim. Many words of this class in the Glosses are evidently learned words, invented to translate the Latin.

DEPONENTFORMS. ACTIVBFORMS. adamgnr, admiror (adamre). noadamrugur, Wb. 16c pret. adruarnraigset, g. sunt sg. 1. pl. 3. 1 admirati, Ml. 888 18. 3. adamrigethar, Wb. 5c 16. innahi adamr~edar,g. quae miratur, M1. 64c 19. pl. 3. adawaigetar, g. mi- rantur, MI. 124a 6. subj. sg. 2. adnamraigther, 91. noli mirari, H1. 56b 39.

1 But 2uae arroneith, M1. 60b 9. From ind-neithins comes a deponential imperative indnite, Wb. 10%21. Otherwise the forma are active. THE DEPOXENT VERB IN IRISH-J. STRACHAN. 469

DEPONENTFORMS. ACTIVEFORYS. pl. 3. coadamraigthar (1.-etar, Ascoli), g. ut mi- rentur, MI. 39b 8. adnamraigetar, g. de- beant mirari, MI. 39b 10.

"adbartaignr,' aversor (cf. adbnrtaig .i. ab adversario, Cormac's Glossary, YBL.). pres. ind. adbartaigther, g. Bver- co adbartaigid, g. ut arer- sg. 2. saris, MI. 44b 31. "s"p: 1 setur, Ml. 36a 11. 3. ni adbartaigedar, g. pret. roadbartaigaet, M1. 26b non aversatur, M1. p1.3. 1 20. 36a 18. pret. sg. 1. roadbartalgsiur, g. fut. annunadbwtaigfeai, g. aversatus sum, MI. sg. 2. 1 te arersato, MI. 488 12. 115& 13, 11'lC 5. 3. niroadbwtaigestar, g. non fuerat aversatus, M1. 55d 1. fut. sg. 1. adbartaigferaa, g. aver- sabor, MI. 37c 12. "adkitchur, detestor (ade'teg). pres. ind. adeilchethar, MI. 50d9; sg. 3. f cf. 122b 12,129c 10. ipv. sg. 2. aditchideaiu, g. de- testare, M1.103a 7b. "adribaigur, resulto.2 pres. ind. adribaigedar, g. resultat, MI. sg. 3. 1 146b3. "aidlicnigur, egeo (aidlieen, egens). an nadnaidlicnigedar, pres. ind. g. nullo egens adiu- sg. 3. I tore, MI. 130R2.

1 The treatment of db under the accent indicates that this is a learned word, probably coined to express mereor. Is it a compound of bertoigim, nbro 2 * On this verb cf. Ascoli, Lexicon Palieo-hibemicum, ccriii. Phil Trans. 1801-2-5. 31 470 THE DEPONENT VERB IN IRISH-J. STRACHAN. DEPONENTFORMS. ACTIVEFORMS. nad adilgnigedar, non eget, Pcr. 56&2. pl. 1. ni adilgnGmar, Wb. 15a 2. 3. aidlignitir (1. aidlfgni- gilir, Asc.), Sg. 4b 10. ai aidimigetar, non egent, Sg. 200&2.

*ainmnigur, nomino (ainm). pret. sg. 3. rondnainmnigestar, no- minavit, Ihil. 17b 9; of. 26b 8, 37b 22.

airigur, percipio (aired ; cf. Asc. Lex. xxvi.). pres. ind. ni conairigursa, non sg. 1. 1 percipio, Wb. 16828. pret. sg. 1. hdre ndnrairigsiur,Wb. 3c 26. ndd rairgsiur, Wb. 3" 27.

sardrigur, appareo (airdrech, ostentum). pres. ind. ardrigitir, g. apparent, pres. ind. ardrigid, g. apparet, pl. 3. 1 Bv. 1" 4. sg. 3. 1 Bcr. 33d 7. pret. sg. 3. roardrigestar, g. ap- paruit, Wb. 280 12.

Sbeoigur, vivifico (hko). pres. ind. heoigidir, Wb. 13d sg. 3. 7. nombeoiysdar, Wb. 19a 20.

*bindip, concors sum (hind). donaib bindigeddar, pres. ind. hi eis quie concordant, PI. 3. sg. 108 9. THE DEPONENT VERB IN IRISH-J. STRACHAN. 471

DEPONENTFORMS. , ACTIVEFORMS. "boltigur, olefacio (bolaod, odor). pres. ind. nibo1tl;getarside bolad, subj. boltigme, g. ohere, nonsentiuntodorem, pl. 1. Acr. 57. pl. 3. I Wb. Iqd 34. "bresminigur, frangor (breisim .i. gciir, -me%- stem). pres. subj. bresrnina$edw, g. fran- sg. 3. 1 gar& M1. 10sd 6. bruthnaigim, ferreo (bruth, fervor). pres. ind. cianudbruthnaigedar, pres. ind. brulhnaigim, g. furo, quamquam fervet, sg. 1. sg. 54b 7. sg. 3. 1 Ml. 1218 15. cairignr, vitupero (eaLre, g. nota). pres. ind. nochairigur, Wb. 1 Id 1. sg. 1: f 2. lases nocairigther, Wb. 10 10. 3. cairigedar, Wb. 25b 9. subj. sg. 2. diacairigther, Wb. lc 10. "cathaigur, pugno (cath, pugna). subj. sg. 3. friscathaigedar, g. re- pres. ind. a cathaigte, g. con- pellare, Nl. 908 6. pl. 3. fligentes, MI. 65" 8. pret. sg. 1. rocathichsiur, Wb. 24a3. Scestaigur, disputo (ceist, quastio). pret. sg. 2. arrucestaigeer, cum disputasti, M1.2d 3. "cichnaigur, strido. pret. sg. 3. ctchnaigietir,' g. stri- derat, Sg. 152b 2. *clnichigur, ludo (chick, ludus). pres. ind. acluichigegedar (1. cki- sg. 3. t chigedar), MI. 122* 10.

I Formed by analogy a~ the absohite form of the 3 sg. of the pret., as were formed later, e.g. to the pres. sec. foad 'he slept,' an absolute foaid, and to the 3 pl. pret. rogabsat, an absolute gabsit. The verb seems based on a noun borrowed from a British dialect with the common analogical change of p to G; cf. 0.W. pipemmu, g. artenas, Mod.W., Bret., and Corn. pib, Corn. piba ' to pipe.' 472 THE DEPONENT VERB IN IRLSH-J. STRACHAN.

DEPONENTFORMS. ACTIVBFORMS. dondi cluchigedar, g. ludenti, M1.1228 12. pl. 3. intan cluichigetar, M1. 129c 20. “cobrigur, auxilior (cohair).

scomadasaigur, apto (cornadas, aptus). pres. ind. comadasaigidir, g. ap- 50”. 3. f tat, Ml. 53c 8. comadassaigidir, g. ac- comodat, Ml. 80d 6. “comaicsigur, approprio (comocus). subj. sg. 3. comaicsigedar,g.adpro- fut. comaicsigjd, 53b 17. priare, 1511. 107b 9. sg.3. JFcomdemnigur,’dominor. subj. sg. 3. comdemnigedar, gl. do- minari, Nl. 17b 1. “cosmailigur, comparo (cosmail). acosmiligmmer, cum pres. ind. eomparamus, Sg. pl. 1. i 21 1% 14. 3. cenodchosmaihgetar, quamquam similes sunt, Sg. 212b 2. pret. sg. 3. roscosmailigestar, com- paravit eas, MI. 55d a. *cruthaigur, formo (crzcth). pres. ind. cruthaigedar, g. plas- mantis, i.e. qui plas- “. 3. mat, M1. 140b 5.

Cf. coinz&~~naeht,g. dominatum, M1. 358 4, gen. -ac?ita, 17h 6. Cf. demin ‘certain,’ demnigina I make certain,’ Or have we a derivative, as Siegfried held, from an Irish stem cognate vith Lat. domitaw ? . Cf. Johansson, IF. iii. 298. THE DEPOIiENT VERB IN IRISH-J. STRACHAN. 473

DEPONENTFORMS. ACTIVEFORMS. pret. pl. 1. rochriitha~semm~r~sg. 98 22. "cuimnigur, reminiscor (cuimne, memoria ; cuimnech, nzemor). cuimnigedar, g. remi- pres. ind. niseentis, i.e. qui sg. 3. reminiscitur, Wb. i 16b 24. *culigur, profano (col). pret. sg. 3. arruculigestar, g. pro- f[an]ante, MI. 6.P 14. cumachtaigim, potior (cumachtach, potens). diacumachtaigther, pres. ind. cia pres. ind. cumachtaigim, g. quo potiris, Sg. potior, Sg. 39b sg. 2. sg. 1. i 20gb 30. 2; cf. 54& 7. honacumachtaigfet, g. quo non sine potituri, Ml. pl. 3. 288 12. cumgigim, ango (cumce). pret. sg. 3. racumgaigsstarsom, g. pres. ind. cumcZgim, Sg. 16b 9. se angiavit, MI. sg. 1. 133&9. *cutrummaigur, exsquo (cutrumme). subj. sg. 3. cochutrummaigidir, g. ut exsequet, Ml. 250 12. pret. sg. 3. rocutrumnaaigestar,ex- equaTit, MI. 55d 3. daingnigim, munio (daingen, firmus). pret. sg. 3. rodaingnigestar, MI. 51d 8. subj. conumdaingnigess, g. ut (me) communias, Ml. 6g. 9. 4gd 13. lame nundundaingnkhfe, g. cum n6s munieris, MI. 78C 6. 474 THE DEPONENT VERB IN IBISH-J. STRACHAR.

DEPONENTFORXS. ACTIVEFORMS. Sdinaignr, dono (ddn, donum, ars). pres. ind. inti naddanaiged w], connachaddnaigfea ' so sg. 3. 17c 7. that he will not give 1 M1. sg. 3. it,' Ml. 96a 7. arldanigfea, g. remunera- turum esse, M1. 112c 4. subj. sg. 2. manidanaigthersu, nisi des, hll. 40b 2. pret. sg. 3. ronddnaigestar, que dedit, M1.96b 9; cf. 97d 17, Wb. 21b 9. dechrigim, differo, disto (dechur, differentia). prrs. ind. dechrigrnirni, g. dis- pres. ind. dechrigim, g. dis- pl. 1. 1 tamus, 351. 117b 9. sg. 1. f pisco, Sg. 177b 1. pl.3. ni dechriget, Pcr. 62a 2. 3. andechrigeddar, g. dis- andechraigte, g. dis- tantia, sg. 46" 9. tantia, Ml. 119d 1 1. ni dechrigetar, g. non fut. nidechrairg]fet, non dis- absistunt, Sg. 155b4. pl. 3. f tabunt, MI. 90b 6. nidechraigetar, Ml. 113a2; cf.Yf. 115b 2.' mbj. sg. 3. mani dechreedar ' unless it be different,' Wb. 12c 46. Qdeligur,distinguo, separo. pres. sg. 3. nodadeligedw (sonus), Sg. 6%19. ndddeligedar, Sg. 28a 15. "demnigur, firmo (demin). pres. ind. dmnigid, Sg. 158b sg. 3. 1 3. pret. sg. 3. radeimnigestar, Wb. pl. 1. demnigmini, Ml. 35b 1. 32c 20. 3. demnigte, gl. quae com- moniunt, Ml. 75a 8; cf. 16a 15.

1 According to Mr. Stokes the MS. has ~to8&chrigetar. THE DEPONENT VERB IN IRISH-J. STRACHAN. 475

DEPONENTFORMS. ACTIVEFORMS. dephthigim, dissideo (debuith, dissensio j debthach, con- tentiosus). pret. pl. 3. rodebthicbetar, g.dis- pres. ind. nidephthiginz, g. non siderant, M1.190 16. sg. 1. 1 dissideo, M1.218 2. pl. 3. debthaigte, qui dissident, Ml. 103d 17. *desimrechtaigur, pro exemplo pono (desimrecht, ex- emplum). pres. sg. 3. desimrechta$edm, Sg. llb 4. Wsruthaignr, deriro (di+sruth, flumen). pres. hd. disruthaigedar,derirat, sg. 3. 1 sg. 198b 3. *dinmmussaigur, superbio (diummus, euperbin; diummsach, superbus). subj. sg. 3. arnadiummussaigedar, g. ne turgescat, M1. 6W 7. *dixnngur,l sum.

sg. 2. 3. naddixnigedar, non est, Wb. 9c 14 ; cf. MI. 2OC 6, 510 16, 55O 10, Sg. 293, etc. pl. 3. dixnigetar, Wb. 12b 7. subj. sg. 3. indhnigedar, g. inesse, Bdl. 19b 14. *dochenelaigar, degenero (dochendl, degener). pret. sg. 1. arrondoichenelaigsiursa, g. degenerans, i.e. cum degeneravi, M1. 44b 36.

1 Cf. digas, editus, M1. 32a 16, 41c 9, 106d 12. air-dim, producta? Mr. Stokes suggest4 that *&'zaigur may be for *di-aicsnigw, from ahits, gen. aicsen ' vision,' literally appareo. 476 THE DEPONENT VERB IN IRISH-J. STRACHAN.

DEPONENTFORMS. Am~wFORMS. "dogailsip, maereo (dogailse, maeror). pres. ind. condogaihegatar, Ml. pl. 3. f 8Tb 17. *dumaigur, exaggero (duma,agger). pret. sg. 3. rodumaigestar, g. ex- . aggeravit, Ml. 55d 3 ; cf. 80b 3, 83d 1. "echtrannaigur, me extraneum reddo (echtrann). pret. pl. 3. roechtrannaigsetar, Ml. 66d 2. %cmailtigur, insolesco (kcmailt, insolens ; CcmaiZte, insolentia).

sg. 3. "Bcndaigur, obtrecto (Ccizdach, detrectatio). pres: ind. acfidigitir, g. detra- pl. 3. f hunt, M1. 126d 16. "corn-Bcnignr, cog0 (Ccen). nl come'icnigedar, non pres. ind. come'icnisset (1. nig- sg. 3. cogit, Sg. 9. 1 61a pl. 3. set, Asc.), Tur. 2d 18. "engraccigur, locum obtineo (cf. in-engraicc, loco, ~g. 30b 16). mo pres. ind. engraicigidir ainm, sg. 200b 10; cf. sg. 1. 19P 5. nengracigedar, Sg. 1 97h2. pl. 3. nodaengraicigetar (rel. ) , sg. 198b 8. huare nengraicigetar, sg. 200h 5. *erbirigur, causor (cf. airbhire 'reproach,' blame,' O'ReilIy). subj. sg. 3. erbirigithir, g. cause- edirigidid (1. wbi- pres. ind. tur, hI1. 35' 6. rigid), g. causa- sg. 3. tur, M1. 908 7. IRE DEPONENT VERB IN IRISH-3. STHACRAN. 477

DEPONENTFORMS. ACTIVEFORMS. "erdarcaigur, oelebro (erdairo, illustris). pres. ind. erdarcaiggtdir, g. con- fut. noterdwcugu6, g. cde- sg. 3. f celebrat, M1. 28b15. sg. 1. f brabo t6, MI. 558 5. 3. erdwcaigf88, MI. 8Db 4. "erladaigur, obedio (alilithe, obediens). erladaigidir, g. obse- pr:. ? 1 quitur, M1. 64d 5. amal nerladaigedar, g. tamquam obsequi- tur, XI. 64d 3. subj. sg. 1. coerladaLg:ar,' g. ad pariendum, M1.106C6. pret. pl. 3. rot'rladt&etar, Wb. 7C 16. "etaigur, amulor (kt). nobkttigetar, g. Bemu- pres. ind. lantur vobis, Wb. PI. 3. i 19d 27. subj. ag. 2. Izone'taigther, g. noli emulari, MI. ~56~17, 39. nietaigthersu, M1. 56b 31. *etarcnaignr, noto (etargne, intellectus).

sg. 3. subj. sg. 3. renzilaetarcnigedar, g. prsenotare, M1. 18c 12. pret. sg. 3. roetarcnaigestar, g. notavit, MI. 32b 5. "etuailngignr, g. indignor (ktualang). innahi atuailngigedar, pres. ind. g. quse indignatur, sg. 3. Ml. 6gd 7.

1 Ascoli remarks 'non b effettivamente se non -gear.' The regular ending in these verbs would be -ger. 478 THE DEPONENT VERB IN IRISH-$. STPACHAN. DEPONENTFORMS. ACTIVEFORXS. pret. sg. 3. wrue'tuailngiata (1. g@star) &a, g. deo indignante, Ml. 62b 22. Sfailtigur, lstor [failte, laetitia). subj. sg. 1. rzzufailtLger, g. letas, pres. id. failtigmi, g. kta- M1. 46b 16. pl. 1. 1 mur, MI. 12gd 7. pl. 3. eonifailtigetar, g. non latentur, Ml. 55b 6. Yaseaignr, vasto, Crdssaigur= "as-ro-fhdssaigur ; irritum facio (fds, fdsach). pres. ind. n~sesraesaigeda,~M1. tz.aheras@ef, ne irrita sg. 3. 1 51b 27. z!::1 faciant, Wb. 30b 10. pret. sg. 3. rofassaigestar, g. ex- hausit, M1. 118b '7. Veidligur, perduro, maneo (feidil). pres. ind. ind. ire;, Ml. ) Izlfidligedw, Wb. 2c 3. pres' 1 sg. 3. PI. 3. subj. pl. I. cofedl@ner,Wb.6a 11. feidliyast~,MI. 21d 4. pet. pl. 1. rufeidligsenzmar, MI. pret. rgeidligset, Ml. 4. 10P 4. pl. 3. 1 10.P fut. afeidzigjd, Ml. 128a 7. sg. 3. ) pl. 3. ni tremnfeidligfeet, M1. 21a 4. Vercaigur, irascor (ferg, fergach). pres. ind. ni fmcaigedar, non sg. 3. f irascitur, Nl. 24b 18. ipv. sg. 2. fercaigthesi, irascere, Ml. 20b 13. Yiachaigur, debeo (fiach, debitum ; f iachach, debens). subj. sg. 3, cia dia$achaigedar, g. ad quem debeat, M1. 44" 3.

1 We should have expected rather nodrassaigedar. THE DEPONENT VERB IN IRISH-J. STRACH.4N. 479

DEPONENTFORMS. ACTIVEFORMS. Siindbadaignr, beatific0 (Jindhadach,’ beatus, MI. 1 14” 7 j cf. 56” 44). indi nodam)nd6adaige- pres. ind. tarsa, g. iecientes pl. 3. i me, M1. 39d 10. Yirianaigur, iustifico (firi‘an, iustus). pres. ind. nodon$rianigedar, sg. 3. 1 Wb. lQa14. firianigedar, Wb. lgb 1. pret. sg. 3. rajrianiqestar, Wb. lgb 13. rondfirianaigestar, h[l. 19d 16. “focridigur, accingo. subj. sg. 2. fotchridigthersu,g.praj- cingere, MI. l0lc 3. 3. focridigedar, g. ac- cingat, Mi. 35O 32. ipv. sg. 2. fochridigthe, g. a- cingere, Ml. 27e 5. Yogrigur, sono (fogur, sonus). pres. ind. fogrigedw, g. sonans, pres. ind. 1fograigte, MI. 42C 2. sg. 3. 1 Wb. 12b 27, 28. pl. 3. %@oilsigur,manifesto (foibe,fo ilsech). pres. ind. foilsiqidir, Ml. P4c 13, pres. ind. foilsigdde, Sg. 200” Bg. 3. f 53” 27. pl. 3. 1 6. nundfoilsigedar, M1. 42b 18. foilsigedar, quod mani- pret. arufdibkset, g. publi- festat, M1. 74“ 13, pl.3.1 cantes, Ml. 748 4. 85” 11. pl. 3. nondafait8igetar, eos, Sg. 71” 7 ; cf. 198* 24.

1 Findbadach=$nd-bithach from $nd ‘white’ and hifh ‘world’; cf. W. gwynfyd ‘happiness,’ and, for the meaning of find, is find ceinbethu, g. bati quorum remises aunt iniquitates, Wb. 2c 2. 480 THE DEPONENT VERB IN IRISH-J. STRACHAN.

DEPONENTFORXS. ACTIVE FORXS. ipv. sg. 2. foilsigthe, g. revela, MI. 56c 2. pret. sg. 3. ronfailsigestar,id mani- fut' fuilsigJd, Ml. 6gd 8. festarit, Wb. 318 9, sg. 3. 1 of. Ill. 51d 14, 103d 11, logb 2, 149 4. pl. 3. eonidrofoilsigsetar, id manifestarunt, Wb. 21c 22, Yollaigur, negligo (fullach, negligens, pl. follig, Wb. 68 22'1. pres. ind. pl. 3. 1follaigeiar, mi29a2b. Yoirbthigur, pedcio, proficio (fuirbthe, perfectus). pres. ind. foir6thigedar, Wb. 4d sg. 3. 1 6. subj. sg. 1. rofoirbthiger, g. con- summavero, Wb. 7% 9. pret. sg. 2. rofoirbthichser, g. pro- ficisti, Ml. 50c 13 ; cf. 43d 17. %fortachtaigur,auxilior (fwtacht, auxilium). pres. ind. nodafortachtalgedm, sg. 3. 1 M1. 25c 5. 'frepthanaignr, medicor (frepad,' g. freptha). pret. sg. 1. cl.rrzlji.e~tanazthiisiur, (1. Ifrepthanaigisiur), g. medicatus, M1. 1038 6. Qgaimigur,hiemo (gaim, hiems). fut. sg. 1. gaimigfer, Wb. 14a 9. SgnLthaigar, assuesco (gna'th, gndthach, solitus). prct. PI. 3. rognathaigsetar, g. ad- sueti sunt, Ml. 34b2.

1 =*frith-bati- ; cf. fTisbensom, g. medetur, M1. 125c 4 ; cf. epaid, Zimmer, Eeltisehe StudieN, i. 122 sq. THE DEPONENT VERB IN IRISH-J. STRACHAN. 481

DEPONENTFORMS. ACTIVEFORYS. *grBschaigur, continuo (grdaach, continu us). aubj. sg. 3. grLachaigidir, g. con- pres. ind. greschaigte, g. fre- tinuet, Ml. 85a 5. pl. 3. quentium, Ml. Wb3. grceachaigfeasa, g. con- it";. ] tinuabo, MI. 122b 10. "guaigur,' mentior (968, falsum). pres. ind. an.quaigedar, g. men- sg. 3. ) tiente, M1. 3Ib 1. pl. 3. guaigitir, mentiuntur, Ml. 31b 1.

Jcilaignr,- multiplico (iZ, multus). ilaigfe, g. multiplicasti, nl hilaigedat-, non mul- fut. ! i.8. multiplicabis, 1\11. tiplicat, Sg. 166a 2. sg'2' /, 81C 5. cid ara Izllaigedar, cur multiplicat, Sg. Wb7. Jcimdaigur,aredundo, affluo (immda, abundans). imdaigidir, g. redun- subj. nonimdaGi, g. polleas, pres. ind. dat, N1. 71C 10; cf. sg. 2. Nl. 88b 6. sg. 3. i 12lb 14, 68b 11. fut. pl. 3. imdaigitir, g. reddun- imd,i.fid, g. uberabitur, sg.3. dant, N1. 70b 5 ; cf. Ml. 81C 11. 39d 2. imdaigetar, g. affluen- tia, hll. 39d 5.

Jcinbothignr, nubo (inhotha, nuptias, Tur. 48). subj. pl. 3. dianidothigetm, Wb. 29 1. %linaigur, inretio (ilz+lin, rete). subj. pl. 3. inlinaigetar, g. inre- tiere, M1. 328 14.

1 In the pass. niruguigtev gnimui dce, g. mentiri nesciam, M1. 51C 14. In an imduig dia, g. deo inminente, MI. 6Bd 18. Ascoli takes imduig to be 3 sg. pret. (cum deus preevaluit) but in these verbs the 3 sg. pret. is regularly deponent. Should we correct to inzduigid, -id last before dP 482 THE DEPONENT VERH IN IRISH-J. STRACHAN,

DEPONENTFORMS. ACTIVZFORMS. "inraiecaigur, dignor (inricc, dignus). subj. sg. 2. rnani inraiceaigther, g. si dedigneris, Y1. 103a 8. isligur, humilio (iseZ, humilis). pres. ind. nomishgur, Wb. 17d sg. 1. 1 22. "lethnaigur, dilato (letha),, latus). pret. sg. 2. rokthnaigser, g. dele- tasti, MI. 50a 14. "lobraigur, Bgroto, languesco, infirm0 (Zobur, infirmus ; lobre, infirmitas). pres. ind. Zohraigedar, g. cum sg. 3. egrotat, Ml. 43d 2 1; cf. 96b 12. pl. 3. Zobrat$etar, g. egres- centium,Nl. 131C6. subj. sg. 3. arna lobraigedar, g. ne languescat, Ml. 7 Ib 6. londaigim, indignor (load). pres. ind. indi londabedar, g. in- Zondaigim, g. aspernor, sg. 3. dignantis, &El.64b 5 j Pcr. 62b 3. 1 cf. 102b 11. pret. sg. 3. rolondaigestar, g. in- dignatione commo- Tit, Ml. 29a 2.

YomiCmaigur, subigo (fo+mdm). subj. sg. 3. fornmama~edar,g. su- bigerit, M1. 113c 6. pret. sg. 3. fosroarnmamigestar (1. fwro-, Stokes), MI. 67b 24.

1 Another compound with doir 'enslaved' is found M1. 931 9, rondoivenaigsm (corr. -mamaigsem, Asc.), 1Ola 8, rondoiiammaigestar according to Mr. Stokes (1. rondoirmammaigcstar, Asc.) . THE DEPONENT VERB IN IRISH-J. STRACHAN. 483

DEPOITEITTFORMS. ACTIVEFORMS. asmecnugur, eradico (es+meccon). pres. ind. coasmecnugursa,' g. ut 90". 1. 1 eradicem, 111. 23 8. 'mescaigur, inebrio (mesc, mesce). fut. sg. 2. mescaQjder, g. ine- briasti, i.e. inebria- bis, Ml. 81c 4. *mencigur, increbesco (menicc). menciqite, g. fre- pret. sg. 3. romincigestar, g.incre- pres. ind. quentiurn, Xl. buit, Ml. 36a 40. pl. 3. 88b 2. Ymindchigur, emendico (mindech, g. tenuis). pres. ind. mindchigitir, g. emen- pl. 3. f dicant, M1. 36C 18. Qmiscsigur, odi (miscuis, odium ; miscsecb, exosus). pret. sg. 3. romiscsigestar, Wb. 4c 16. Qm(iithaigur, emollio (mdith, mollis). subj. sg. 3. eomoithaigidir, g. ut emolliat, Ml. 131a 9. Qmothdgur, stupeo (moth, stupor, M1. 68b 9). prrs. ind. mothaigedar, g. stu- comothaigid, g. ut stu- sg. 3. 1 pentis, MI. 26b 9. :::: f peat, Hl. 25C 6. %nrechtnigur, vario (mrecht, varius). pret. sg. 3. ciarud mrechtnigestar, g. variavit, Ml. 123b 12. *oenaigur, unifico (oen, U~UB). pres. ind. fobith koenaigedar, Sg. sg. 3. 1 1128 4.

1 Confusion between -ur and -ar is found in the glosses in the 6 fut. and subj. But I have no examples of -ur in the subj. as in later Irish. The form here is, then, probably indicative, ' so tr. at I do root out.' 484 THE DEPONENT VERB IN IRISH-J. STRACHAX. DEPONENTFORMS. ACTIVEFORMS. Yordaigur, ordino (or& ordo). pret. sg. 3. rodordigestar, Wb., 68 3; cf. 6a 4. "reidigur (rkid, planus). niredigedw, g. nihil commodat,M1.24d22. "sainigur, diversus sum (din, diversus). pres- sainigedar, Sg. 41b 9. sg. 3. f %hithraignr, hboro (ediihar, labor). pret. sg. 3. rusaithraigestar, Nl. pres. ind. ' huare saithraigte, 92c 5. pl. 3. ,I xi. 12s a. *sririgur, viol0 (sdr). pret. sg. 3. rusarigestar,M1.71b 14. pret. pl. 3. rdsa'richset, Wb. 1". Sscithigur, defetiscor (scith). fut. sg. 1. niconscithigfar, g. non defetiscar, Acr. 58. %6migur, attenuo (skim). subj. sg. 3. cosemigidzi., g. ut at- tenuet, MI. 44d 9. "senaigur, inveterasco (sen). subj. sg. 3. maienaigidir, g. si inveteraverit, Sg. 151a 1. "silaigur, sero (ail, semen). lase nosilaigid, g. subj. sg. 1. nosilaiger,'g. exserere, pres. ind. adserendo, Ml. M1. 23c 28. , sg. 3. 115h 11. %ochenelajgur, nobilito (sochend). subj. sg. 3. cosochenelaigidir, g. ut nuhilitet, MI. 138C 4. "sonartnaigur, valeo, convalesco (sonairt). pres. ind. nisonartnal:gedar, non sg. 3. ) valet, M1. 14C 4.

A literal translation of Lat. ~ero' sow.' THE DEPONENT VERB IN IRISH-J. STRACHAN. 485

DEPONENTFOEMS. ACTIVEFOBMS. subj. sg. 3. arnaaonwhaigedw, g. ne convalescat, M1. . 28d 9. pret. sg. 3. arrroaonwtnaigestar, g. convalescens, MI. 49b 4. *subaigur, gaudeo (&a, gaudium ; subach, lstus). auba(qit, g. gaudent, M1. sg. 3. ::. 1 61C 18. *suidigur, pono (wide, sedes, eessio). pres. ind. osuidigedar, g. pin- suid*$th, ponet, MI. sg. 3. ) guit, Sg. 498 9. ~:.~3. f 46c 20. fondasuidigedw, easup- ponit, Sg. 161b 5. pl. 3. suidigitir, ponunt, Ml. 940 3. pret. sg. 1. rusuidigsiuraa, g. statui, M1. 5gb 2. 2. rosudigsmw, statuisti, Ml. 1218 12. 3. roauidigestar, posuit, Wb. 128 30; cf. M1.460 20, 630 10, 1308 7. PI. 3. fMYU8Uk?ig86tW, g. SUP posuerunt, Wb. 7b5. "suthaigur, fructuosum facio (suth, fetus ; suthach, fructuosus). pret. sg. 2. rosuthchaigser, g. foe- tasti, Ml. 8Ib 9. ittaitnigiur (taitnech). pret. sg. 2. rotaitnigsersu, g. pla- catus es, MI. 105e 7. ittessaigur, tepefacio (tess). nade:tesaigedar,' g. pres. ind. quos nullos tepe- sg. 3. facit, M1. 94b 2 1. Cf. Ascoli's note on the passage. According to Mr. Stokes the MS. looks like nudeltesaigedar, a scribal error for nadateaa%gtdur. Phil. Trans. 1891-2-3. 32 486 THE DEPONENT VERB IN IRISH+. STRACRAN. DBPONENTFOXME. AGTIYBFOXME. yret. sg. 3. conruteesa@eatar, g. concaluit m6, Ml. 5g8 16. otoirthignr (toirthech, fructuosus). anaal toirtha'gedar, pres. ind. . . . ut fructum fert, sg. 3. t Wb. 26c 3. *toltanaignr, placere (toltanaoh, g. bene placitus). pret. sg. 3. rotoltanaigeatar, g. quod Caessari . . . placitum, Sg. 7b 10. toridgiur, fido (torianich, g. fidentes, Wb. 14b 27, cf. Ml. 18). pres. ind. nothwisnigiur, g.fiden- pl. 3. taraimigte, g. fidere, Ml. sg. 1. 1 tem, M1. 1264 19. 39b 9. pret. sg. 3. rotorasna@estar,g. con- fidendo, MI. 106b 8. qrebrigur, continuo.

ipv. sg. 2. trebrigtha, g. perpetuir, MI. 88d 10. trdnaigim, (trCn, fortis). pres. ind. immetrenaigedar, M1. ind. imtdnigim, g. mando, sg. 3. f 62C5. sg.1. f Sg. 146b 4. Yuailngignr, dignor (tualaing). nuntuailngigedar, g. pres. ind. nos dignatur, eg. 3. . . . MI. 146b 2. pret. sg. 3. ni rothuailngigestar, g. non est dignatus, MI. 16b 12. Jcuaibrigur, fallo, frustra habeo (uaibrrch, fallax, inanis). pret. sg. 3. runuaa'bra'gestar,g.quos profanaverat, M1. 73b 10. THE DEPONENT VERB IN IRISH+. STRACHAN. 487

DEPONENTFORMS. A~IYBFORMS. Onatigur, raresco (uate, singularis). pres. ind. kuatigitti; g. rare& pl. 3. 1 cunt, MI. 338 15. *iUaignr, vireo (hr, viridis). subj. sg. 3. nuraigedm, g. virere, M1. 15b 14. *ntmallaigur, nuto, vacillo (utmall, mobilis, instabilis). subj. pl. 3. dhinnadnutmahgetar, g.utrumnon nutent, Acr. 45. utntallaigctar, g. vacil- lare, Acr. 82.

The following verbs in -aigim show no deponent forms, with the exception of the 2 ipv. in -the, which is the only ending that these verbs have in this form. Many of them are found only in the 1 pg. pres. ind., a form in which the active ending is predominant in the Glosses.

Oaoarbaigim, aspero (acarb, asper), acarbaigte, g. aspernantia (1. asperantia), MI. 87b 14; hctegim, g. acesco, Pcr. 578 3; alterno (uile), digid, Sg. I668 2 ; *airigh (airech, primas), noairiget, g. primi ussi sunt, Ml. 125d 4; *aithirgigim, me pamitet (aithirge) lase ra-aithirgigte, g. penitendo, MI. 105d 11; hlgenaigim, lento (Mgen), Sg. 14P 1, cf. Pcr. 56b 3, Cilgenigd, Wb. 4a 2 ; *arsadaigim, veterasco (arsaid), arsadaigjth, veterescet, M1. 6gb 4; athrigim, g. patrisso (athir), Sg. llb 7 ; baithaigim, g. futio (baith), Sg. 56b 5, cf. 56b 6; "bertaigim, vibro, robertaigaet, g. vibraverunt, Ml. 26d 5 ; cniligim, g. prosto (czdihch, g. prostibulum), Sg. 53&18 ; *aitsachtaigim, furo (ddeacht, dhachtach), dasachtaigte, g. furientium, Acr. 50 ; "dengaigim, pot0 (lleq), dmgaigjt, g. potabunt, M1. 30c 18 ; "dimagim, celero (dian), dianaigthe, g. celera, M1. 49 9; *dilmctinaigim, VBCO (dilmain, g. expeditum, M1. 8Ib 7, is diZmain, licet), rondihainaigaet, g. vacasse, M1. 768 8 ; %lithrigim, exulo (aiZzthrech. peregrinator religiosus), elitkigmi, g. exulamua, MI. 46O 22 ; *6trmmaigim, lev0 (klromm), rodtrummaigset, Ml. 113c 8, ktrummaigjd, g. sub- 458 THE DEPONENT VERB IN IRISH-J. STRACHAH. levabit, MI. 89c 9 j failligim, g. sarpo, Sg. 183" 1 ; feecrigh,' vesperasco (fescw, vesper), g. obsero, Sg. 146b 7 ; ?feuchraigim (jhchuir), rofuchraigsst, g. efferati sunt, M1. 1148 6 ; %ofin- bnanaigim, vindemio (fln+ biiain, messis), cofinhuanaigit, g. vindimiant, Ml. 1028 12 ; fliuchaigim, g. lippio (giuch, madidus), Sg. lob 3; foirmtigim, invideo (foimntech, invidus), ndd foirm- tigimse, g. non invideo, Acr. 33; gaigim, g. quirito (gdi, hasta), Sg. 144b 2, Pcr. 56b 2; gortigim, g. sallo (goid, g. acidus), Sg. 1878 6; *hilligim, tutor (inill, tutus), inilligN, g. tutabitur, Ml. 128c 8 ; %tledaigim, insidior (intlide, insidiosus ; intle, in- sidiae), intlsdaigte, g. insidientium, Ml, 390 25 ; intonnaigim, g. inundo (inftonn, unda), Sg. 14g8 2; *lainnigh, avidus sum (lainn, acer, avidus), lainnigte, qui inhiant, M1. 716 4 ; Smadaigim, frustror (madach, cassus), rusnzadaigset fesin, g. ipsos frustrata sunt, MI. 4a8 I, ni rumadaigset, 488 2; mhnaigirn, hio (mdn, hiatus, rictus), menaigte, g. inchiare, Y1. 71d 4 ; Jcmesraigim, tempero, mani mesraigea, MI. 460 15 ; mfinigim, g. mingo (mtn, unna), Sg. 1748 2; *nuallaigirn (nuall), nu[a]llaigem. g. ploremus, Nl. 114d 3 ; *ollaigim, amplico (oll, magnus), ollaigthe, g. amplici, Ml. 700 7; rhchtaigim, g. lego (recht, lex), Sg. 60b 16; conaearcagim, g. condilector (serc, amor), Wb. 3d 9; *setaigim, vior (sit, via), innani skta@te, g. viantium, MI. 82d 4; %sl&naigim,sldn (salrus), sldnaigthe, g. osanna, M1. 25b 2; sulbi[righ], eloquens sum (sulbair, eloquens), Wb. 12d 10 ; "tirmaigim, sicco (tlrim, siccus), cothirmaigid, g. ut exsiccet, MI. 44d 8 ; tosligachtaigimse, g. pendo, MI. 7ga 9, cf. 78e 3 ; trebairigim, g. sapio (trebair, sapiens, callidus), Sg. 146" 16, cf. Pcr. 57" 6; nalligim, g. arrogo (uallach, superbus), Sg. 22b 2 ; Yoriraisligim (Gasal, altus), ni foruwisligemni, g. non auperextendimus nos, Wb. 17b 17.

I The glossator connects obsero with sera ; cf. Sg. 183b 3. In Per. 57b 2 obsero is rightly explained by frisdheim. THE DEPONENT VERB IN IRISH-J. STRACHAN. 489

2. THE IEISKHYMNS.‘ DEPONENTFORMS. ACTITE FOEMS. clainiur. subj. sg. 3. rohchloathar,2vii. 61. cuiriur. to-cuiriw,5 vii. 42. pret. pl. 3. fwdacoraatar, v. 66. fetar. sg. 3. ftir, vi. 7. *gailliur. perf. sg. 3. gknair, ii. 1, 68. adglhdur. pret. sg. 3. adgkidastar, ii. 48. labmr. subj. Bg. 3. rolmlabrathar, Vii. 59. moiniur. pres. ind. admuammar, iii. 1 ;cf, pl. 3. f v. 98, 99. perf. sg. 3. romknair, u. 67.

1 Btokes, GoideZica,2 121 sq. ; Tripartite Zife of Patrick, cxi., CG.,48, 404, 426 ; Windisch, Irische Texts, i. 1 sq. ; Zimmer, Keltische Stzclzien, i. 5 sq., 5. 160 sq. ; Wtschr. f. deutsches Alterthum, xxxv. 7%; Thurneysen, Rev. CeZt. vi. 326 sq. These are traditionally ascribed to several authors, from St. Patrick’s time down to the eighth century. For one of them, Fiacc’s hymn (ii.), Thurneysen has shown (2.c.) that its language is later than the language of the glosses. He points to asbert for asrubart, dobert for &rat&, and the absolute forms of the preterite ya6ais, etc. Such an absolute leicsi is found in a later addition to the Milan gl0SSeS ; the formation is common in Tirechan’s notea in the Book of Armagh; it probably aroae from the suffixation of a pronoun, e.g. cili-s from (ro-)di. Another new formation in this hymn is the absolute secondary present foaid ‘he used INsleep’ to foud. This is clearly a new analogical formation, like the absolute gadsit to royabsat. The language of the other hymns is of the game kind; they may be put, for the most art at least, in the latter part of the ei hth or in the ninth cent &. Patrick’s hymn (vii.) is, however, probatly considerably older ; if Sxes (Trip. Life, ci.) be right, it falls before the Milan glosses. The verbal forms in it are too few to furnish a su5cient basis for corn anson. In the Liber Hymnorum is added the Amra Choluimb Chille, but tL 18 a document of a different sort, written in a purposely obscure and affected style, and evidently not nearly so old as it protesses to be; cf. the forms CZWS 5, sexus 67, cuillsiw 102, cluid8ius 119, ek., a’esestar 120 (=O. Ir. dessid), congein 128, atqaill 123, dorumeoen 108, Gvthier 68. Rawl. B. 612 (Goidelica, 153) ha8 the corruption rodomltcilledar. 3 This form is preserved in Rawl. L. Hy. haa to-cuirius. Apart from other considerations the sense requires a present. 490 THE DEPONENT VERB IN ZRISH-J. RTRACHAN.

DEPONENTFORMS. ACTIVE FORMS. -tlUChW. ipv. sg. 2. atla@he, ii. 49. pret. sg. 3. dotlzc[ch]estw, v. 47. duthraccar. s. fut. sg.3. midhthrastar,l vii. 39.

The dep. subj. of -ciu is found in nom&c&zr, vii. 60; the pret. sia[ss]air, v. 1. Deponent, 3 sg. 8. pret. sknasfar, v. 45 (but sknais, v. 39) ; rodglinnestar, v. 45, noconmillsstar, v. 46; aridrala~tar,~ii. 47 ; cf. v. 75. Isolated deponent forms are timadar3 (1. temadm) ‘may he protect,’ i. 2 ; ronfeladar ‘may he cover us,’ i. 1 ; rontolomarP ‘may we please him,’ i. 35; rodomscdtadar, vii. 48, Rawl. (immimrorda, L. Hy.).

From -ag-, -ig- verbs. Deponent, nochsnagur, v. 105 ; cathaigestar, iii. 3. Active, samaiges, ii. 55 ; tathich, v. 57.

3. CORUC’S GLOSSARY.~ In the &st class of verbs the deponent forms are well preserved.

R. has the prw. midhthracair. In ii. 47 Stokes translates ‘set him in motion’; in v. 75 doubtfully by ‘ wrought ’ (connecting it with lirainr ‘ throw,’ ‘ putt ’T) . The nearest parallel I have noted is conarlastar it& nu dorsaide, Windisch, Wh. 374b. A deponent ipv. is found in Cormac’s Glossary emhde (=eimide, Laws, iv. 32. 23); cf. tmathar .i. ditnither, H. 3. 18, p. 651 (Stokes). 4 Of. ratholathur ‘pleased,’ LL. 1718 3. The Leabhhar Bream version of this work (Br.), with the fragment from the Book of Leinster (I,.), has been published by Stokes in his Three Irish Glossaries. In the Transactions of the Philological Society, 1891-2-3, 149 sq., he has published the fragment contained in Land 610, Bodleian Libra (B.), with readings from the complete copyin the Yellow Book of Lecan (?), and a fragment in Harl. 5280, British Museum. Through the kindness of Mr. Stokes I have been enabled to use his transcript of the Book of Lecan for the earlier part of the Glossary not included in the latter publication. The old forms have been much better preserved in Y. than in Br., 85 a few examples will show:-doimurg, Y. ; ttmnirgaas, Br. 8.v. anforbracht ; timchell, Y. : timchelles, Br. 8.7. Coire Brecain : doe, Y. ; dmt, Br. 8.v. deach. With regard to the date of the work, it is traditionally ascribed to the King-bishop Cormac of Cashel t903. Stokes, Three lriah Glossaries xviii., thinks that the greater part of it was written, “if not in the time of Cormac, at least within a century or so after his death.” In his edition of the Bodleian fragment he now holds that its language proves that it was written not much before the eleventh THE DEPONENT VERB IN IRISH+. STRACHAN. 491

DEPONENTFoaars. Amrw FOBXS. Clllininr. ipv. sg. 2. cluinnte, S.V. auchaide, Y. cuiriur. pres. ind. sg. 3. docuiredar, docuirethar.’ a pret. sg. 3. CEOcwuutair, happened, S.V. prull. fetar. 8. subj. sg. 3. nhz$astar, S.V. Pam. ipv. sg. 2. fintce, S.V. oro. jnnta, S.V. emhde. *gainiur. pres. bd. sg. 3. genither is born,’ 8.v. Buanand, gein ligainethar, L. (genuss, Br.) ‘a creature which is born,’ S.V. tura$in, genethw, Y. (genes no genit[h]er, Br.), a.v. diuthach. pl. 3. geniter, S.V. Emuin. perf. s.g. roghenair, 8.v. Cormac. ad-glhdur. pres. ind. sg. 3. nitaicelladar does not address thee,’ B. (nitaicelfs ‘ will not address thee,’ Br.), S.V. pll. ipv. sg. 2. atomglaiteee, B. (nomaceZZslsi, Br.) address me,’ ib., nahacaldai, B. ib. century. Zinimer, Nennius Vindicatus 89 ; cf. Zeitschrift fir Deutnches Alter- $hum 118-120, holds the view that the work was written by Cormac and “erfuhr mter einem Nachfolger Cormac’s (Brian) zwischen 1001 und 1014 eine Neuausgabe, in der es abgesehen von gelegentlichen Zusatzen ad una gekomen ist.” I hope it will become clear in the course of this investigation that the ori ‘nal work cannot have been composed later than, at the latest, in the first hafof the tenth century, and that any new edition from which our MSS. may have come, must have preserved very faithfully the old forms. Moderni- zahon of the language is certainly found in the LeaQhar Breacc version, and, beside the modem form, the old form is often given aa a variant, e.g. airises no arsiaedar, Br. 8.v. cuisnit: arsisedar, P. ; conmaid no coltma, Br. 8.v. imbaa fmosnad:eoncna cett., but how far back this text goes I do not see any means of det%~&ing. In other respects, too, Br. differs from the other text. It is to he noted that in the Irish interpretations of Latin phraees we find late forms, as 8.v. apstal .i. ab po8dulo .i. ad domcnutn posdulo .i. tdchurtmm, where, corresponding to the language of the bulk of the work, we should expect doehuiriur. Here we can hardly avoid the conclusion that we have to deal with later additions. 1 doeuivedar (doezlirethar, Br.) cachai ituaimm aroli, 8.v. Coire Brecain ; docuiredar iaram forsin choire ‘lights upon the cauldron,’ ib. Y. ; bhlcdocacirethar for aigid duine ‘a swelling which comes on the face of a man,’ 8.v. ferb, Br. ; Y. haa docuirethr indt4ine fora gruadaobh ‘which a man geta upon his cheeks’ ; tocuirethar, LL. ; docuwethar, P., springs (with ellipsis of bedg, of which we shall have more exam h),6.v. taurtlhat. B.V. cartct, Y. gas arumiretar, but aa cosa is norn. a paw. is required, fomcurt[h]ar, Stokes after Br. 492 THE DEPONENT VERB IN IRISH-J. STRACHAN.

DEPONENTFoms. ACTIVEFORWJ. pret. sg. 3. cein cotnaicleatar, S.V. bthech. adgladaatar, S.V. prull. miainr. pres. ind. sg. 1. fwtomidiur (in quotation), S.V. laith. acZmidiw (in quotation), S.V. mann. perf. sg. 3. romidair, romidir, S.V. anair. dorumidir, e.v. Lith. 8. fut. sg. 2. mair (in quotation), S.V. eegamk. siseinr. pres. ind. sg. 3. araieedar, P. (airiaea, Br.) rests on,’ S.V. aurea. araieedar, P. (airbee no ar&edar, Br.), S.V. cuienit.‘ areieedar, S.V. defuach, P. -tlnChnr. pres. ind. sg. 1. atlocor, S.V. arco, dotluchur, Y. (tothlugur, Br.), ib. In addition to these may be mentioned a form, muinethar? found in quotation. S.V. fmmh I take aechid from aechim ‘say’: aechid fo eeinbreith fhin caingin frecnairc ‘ he speaks according to an old judgment in the present case.’ Deponent perfect forms are found in docoemnacair ‘happened,’ s.~.orc and immoacoemarcair ‘ asked them,’ s.~.przcll. From verbs originally deponent active forms are found in

nilaha reil (labrwr) Ihe does not speak clearly,’ B.V. brinda, and in ammolts ‘ when he praises ’ (rnolur), S.V. fli, P. (amolad, Br.). From both of these verbs, as we have seen, active forms are found in the Milan Glosses. From -ag-, -+- verbs. Deponent forms are found in ga1zltgedar3 boils,’ Y., 8.r.

1 In the same gloea Br. has for the sg. foraanairisethar in ddlaige, Y., the pl. foraanairisetar in ddlaigi. The sg. suits the latter part of the glow better. 2 S.V. ebrm:ebrh imamuinethar nteivg .i. imatimcella meirg 7 immmith ‘which rust SUITOW& and eats.’ Br. has the obviously inferior reading ebrdn imamuintear (.i. imatirnchelta) “krg 7 ima[n]ith. Mr. Stokes, taking ima- timchella 88 the translation of immmuinethar, suggests that rnuinethar meam oe8’ qmen Urkeltischer Spachschatz 208; cf. diotomwa, g. Wati, LU. 67a 25. If this be so we may compare with muinethnr in form fuinethar in the artificial language of the Dialogue of the Two Sages, ijuinethar gd .i. ife8dgend g6 ‘in which lying 8ef8,’ LL. 186b 43. But whether imatimchelL is intended to be the exact explanation of immmuinethar ia rendered doubtful by immelois Zuinethar .i. immatimchella suilse ‘which light surrounds,’ E.V. laea. 3 fa cmmiless igaluigedar coire mbis for tein ‘ aa a cauldron bo& that is on the fie.’ THE DEPONENT VERB IN IRISH-. STRACHAK, 493

Coire Brecain, mthigedar ‘ that dignifies ’ (in quot.), S.V. laith, connoenaigedm,’ S.V. Turigen, Y., rorathaigeatar, S.V. Zetkch. Active forms in roainmnigeet, S.V. ~88,oumachlmgsizlm, P. (nrmachta@sina, Br.), S.V. diancecht, deockaigas, Y., S.V. anair, roiuidigset, 8.v. Emuin, Iloszlidigset, P. (rosuidestw, Br.), 8.v. gall, tirmaGe8, Y., 8.v. croioend, urdarcaigim, S.V. ceZehad.

4. OLD HEROICTEXTS. Budacht Morainn (LL. 2938294b ”). DBPONBNTFORMS. ACTIVEFORMS. *b~ur,roar. pres. ind. sg. 3. hiredar, 294b 2. *ciallnr. pres. ind. sg. 3. aruscialladar,4 293b 43.

1 Br. has the passive form comoenaigter, but an original deponent form is more likely to have been corrupted than an original passive. L. and Br. have a different text. 2 Most of these texta are taken from the Leabhar na h- 77tdhi-i (LU.) and the Book of Leinster (LL.) ; one or two come from the Yellow Book of Lecan (Lc.), which, though of later date, contains ancient texts carefully copied, in which the old form8 are well preserved. I should have liked to draw upon Lc. to a greater extent, but the long delay in the publication of the facsimile has made that impossible. The probable antiquity of 8ome of these tslea has been discussed by Zimmer, KZ. xxviii. 426-444, Zeitschrift f. deutsches Alterthum, xxxii. 229-239, xxxiii. 129 sq., xxxv. 1 sq., and by Pflngk-Hartmg, Rev. Celt. xiii. 170 B with very different conclusions. It is vain to attempt to fix the date at wkch these texts assumed practical1 their present form until the language in which they are composed has been hly investigated, of course with due consideration of any historical or archsological evidence. The fact that the deponent flourishes here will help to fix an inferior limit. The Tdin Bd Cwaihige, Toga2 Bruidw Dd Derga, Tochmarc Etciine, Serglige Concutaind, and Pkd Bricrend have been analysed by Zimmer, EZ. xxviii. Zimmer’s composition theory bas been acutely criticised for the Toga1 Bruidne Da Derga, by Nettlau, Rev. Celt. xu., xiii., and for the Win, Rev. Celt. xiv., xv. In the followin lists the addition of (Y. signifies that the form is found in veme; of (r.) t%at it occurs in one of those difficult rhythmical declamations which are so indicated on the side of the MS. 8 Of this annoient and difficult text I have a good version from the Yellow (the Y. text I copied in haste I had no time to decipher, but from R.I.A. 23, N. 27 (A.) (writt~napparently in 1713, with a copious interlinear loss) and R.I.A. 23, N. 10 (B.) (the two texta are almost identical), and ab%reviated copies from LL. 346 (Lz.), and R.I.A. 23, N. 27 (AX.), written in 1714. The language of the text is very ancient; among other things we still find initid ml for later bl, 88 the alliteration shows-is tre fir flatha messrada mdra for fedaib ata manna mil& mlasaalgter: here L. has Blassargter, La. bkziaaetar, but Y. has kbk~i88etar,similarly H., lizblasethar, B., mblaisighther, A. 4 The meaning is not clear to me. H. has arieciallntnr, AS. aris ciallathar; the word does not appear in A., B., or Lz. 494 THE DEPONENT VERB IN IRISH-J. STBACHAN.

DEPONJXNTFORMS. Acnva FOBMS. con-6im, preserve. subj. sg. 3. conoadar, 293b 2 (conofasom, H., Y.). fut. sg. 3. cotnofadar, P., H. ; cotnofathar, B. j cotnofaither, A. ; cotndba, L. 2938 2. cotnofadar, Y. : codanodfhathar, 8'. ; cotndba, L. 294b 11 ; cotnoaha, 294b21 (not in other texts). cuirinr. pres. ind. sg. 3. docuredar, La. 346b 5, Y., H. subj. sg. 3. coradar, 2948 1 ; &vrathm, P., H. ; ni ouirethar, 8.' fetar. s. subj. sg. 3. jastar, A., B. follnur. pres. ind. sg. 3. &tlna&ar, 294b 8 ; foltnathar, Y., H. *folamur. subj. sg. 2. arafolnzaider,2 (?), 2938 44. midinr. pres. ind. sg. 1. atamidiur, 294b 15.S s. fut. Bg. 3. admestlvr, 23g8 24," etc. From -du subj. iarmamdecedcvr, Y., H. (=iarmodecai, L. 293b 42, similarly A., Az, B.). Active forms of deponent verbs are found in nosseichfe5 will follow,' L2.3468 50, and ctuines, Y.,B H. Verbs in -ag-, -@-. cotome'cnigedar compels me,' 294b 16 (cotameignightllLsr,A2.), fossaigedar, 293b 22 (fosaiges, Y., H.), maithighder, Y., mathigdir, H. (leg. mathigedar=mathigcs, L. 2948 21)) humti- gethar, B. (huaisligther, A.), tathigedar, H.

1 Under the influence of the indicative, cf. p. 9, note. 2 So arufolmither, B., arabfolmzcighter fmach with the gloss 7 follamhnaigh TlOfoZZamhnaighthW an@ dghsa, A. ; adfoZZnathar with different text Y., H. J The same phrase is found in A., Aa., B., but in a different connection. 4 admestar ddle dhleman ' he shall .udge the creatures of the Creator.' The glossator ~eemsto have misunderstood the passage ; he interprets mstair duile a dhuilmnhuin 'the creature shall judge its Creator.' In 1. 26 L. has wrongly the nom. takm, so A. admestar ma toirthzcibh takmh, but B. atmestar ma lantorath tulmain. 5 This text shows other later forms. It is curious how sechur has vanished in the profane literature. 6 Fobuan bithsuthain sir$r-thernan suithe chines comad mo chose iarmotha suim in passage corresponding to cluined mo JOSCisrmothd swd, L. 2938 47. TEE DEPONENT VERB IN IRISH-J. STRACRAN. 495

T6in B6 Edich (LL. 2488-252b). DBPONENTFOXMS. ACTIVEFoam. 6gllr. pres. ind. sg. 1. attagzcr, 24gb 41. clnininr. pres. ind. sg. 3. rochlzrilzsthar, 250b 37. cuiriur. pres. ind. sg. 3. doczcirethar, 2508 47. pret. sg. 3. docorartar(happened), 248a 24. co cwastar, 25Qb49. adglitdnr. pres. ind. sg. 3. adgladar,’ 249a 52. pl. 3. hnmoanaeaillet, 2508 6. subj. pl. 1. conacddam, 2528 4. -ciu. subj. pl. 1. conaccaw, 2508 14; cf. 19. 3. cmdaccatat-, 251a 17. Verbs in -up,-ig-. Deponent-roair~yestw, 2508 27, naddnairigmer, 24gb 12.

Sc6l Muicee Maic Datho (LL. 11lb-1 14&). adglhdur. subj. sg. 1. cwotwilliur, l12b 47. esnr. subj. sg. 2. cencone8sara,J 112a 23. Iabrr. subj. sg. 3. cenco Zabradw (v.), 1128 26. Deponent perf. nh darnair (Y.) ‘did not grant,’ 1148 32.

1 It would be easy to correct to adgladadar, but the same form is found elsewhere, e adngluer, M1. 53~17, and the question might be raised whether we have not ‘sadation. Un the other hand adgladudar is found, M1. 1158 6, LlJ. 120* 14. 2 This ancient tale has been edited by Windisch, Irisehe Tezte, i. 93 sp., with variants from H. 3, 18, T.C.D. and Harleian, 5280 Brit. Mus. s On such subjunctive forms see Zimmer, KZ. xxviii. 343 sq. 496 THE DEPONENT VERB IN IRISH-J. STRACHAN.

Verbs in -ag-. Deponent nombertaigedar he shakes himself,' 112* 43, rodmberlaigedar,' I 13b 11, rmbertahjwtar, 113b 7. Active rofhua&ttaa&, 11zb 13.

Longes Mac n-Usnig (LL. 25Yb-261b)? DEPONENTFORMS. ACTIVEFORMS. ad-@r. pres. ind. sg. 3. adna+ethur3 (v.), 25gb 25. beccim. pret. sg. 3. beeeestar (v.), 25gb 32, 33. Sburiur. pres. ind. sg. 3. bhrithar (burethw, Lc.), 25gb 37. clnininr. pres. ind. sg. 3. cluinethar4 (v.), 26gb 24. ipv. sg. 2. cluintise (v.), 25gb 49. Sderdrur, roar. pres. ind. sg. 3. derhethar (v.), 259' 23. pret. sg. 3. roderdrestar, 25gb 43. Sgaininr. perf. sg. 3. rogenair, 25gb 45. eamlur. pres ind. sg. 1. samlamdr (v.), 25gb 34.

Verbs in -ag-. Deponent erechtna&edar (v.), 25gb 26.

1 Better nodbertaigedar, Harl. 2 Edited by Windisch, Irische Z'ezte, i. 59 sp. with variants from the Yellow Book of Lecan, and Egerbon, 1782 Brit. Mns. 3 mdr n-uath adnaigethar mo chride crechtnaigedar ci wid ' great horror which my heart fear^ that the steel (?) wounds.' E,o. has the passive forms ataanaigthar and wechtnaigthar, a common kind of error. Lc. has the deponent forms, but creehti-aaiyegethar. 4 etuasaib cluilzethar ylaim. We should, perhaps, read cllailzetar, and translate 4 ears which hear.' 5 Eg. has introduced the active sanaluilnse 'I compare.' Deponent forms of this verb are rare in these texts except in the ipv. sarnailte. An example is found in a poem, LL. 145b, fritotwmkor ' 1 compare thee,' ll. 4, 8, by the side of frtssamlarm, 3, 7, 9, 10. .THE DEPONENT VERB IN IRISH-J. STRACHAN. 497

TBin B6 Regamain.’ DEPONENTFORMS. ACTIVEFoars. adhgur. pres ind. sg. 1. adagar ‘Ifear,’ 37, 48. pl. 1. atagumar, Eg. 42. fetar. sg. 1. rofhdar, 15 (rofetamur, Eg.). adghidnr . pres. ind. pl. 3. adgladatar, Eg. 39. immusnaoallatar, 12. subj. PI. 1. oonarladmar, 42. a. fut. pl. 3. adglaaemwni1249. con-midiur. 8. fut pl. 1. conme~amw,~Eg. 38. siesiur. pres. ind. sg. 3. imasisedar,‘ 57. pl. 1. fobeieemarni ‘we stand by you,’ 42, 53. TBin B6 Regamna.6 cluiniur. subj. ] chinsm, 40 (cluinizcnt, pl. 1. Eg.). cuiriur. pret. sg. 3. conid coruetair, 4. adglhdur. adyladaithersu, 2 7 pres. ind. (Eg. corrupts into atyla- sg. a. f daig tbrsu). Edited b Windisch, Irische Tezte, ii. 2, 224 sq. from the Yellow Book of Lecan, and Jgerton, 1782. 2 In Lc. 41 OCCUTS conarlasar inni, which, as the corresponding Eg. teat, connarlaidrd inna macco ‘ that ye may address the vouths,’ shows, must mean ‘that ye may address him.’ Arlasar for ar7drd IS an analogical formation to arzsamar arhisatar. Cf. athgenair ‘ye recognise,’ lnsche T&e, ii. 1, 176. This form of the 2 pres. pl. did not succeed in establishing itself, -baw prevailing instead. 9 This form supports my view of coimmestar, M1. 1278 19. Windiach, as I see now, also suggests connection with commus ‘ power.’ Lo. has conisamar ; conicelm is regularly active. 4 imasieedar doib ‘sie bleiben zusammen stehen,’ an impersonal we of the deponent. Cf. imnzmasinithar dbib, LU. 608 3. Edited by Windisch, Irische Zezte, ii. 2, 239 sq., from the Yellow Book of Lecan and Egerton, 1782. 498 THE DEPONENT VERB IN IRISH-J. STRAGHAW.

DXPONBNTFOBMS. ACTIVEFORMS. adgldithersu 34(atgla- dmgthersu, Eg.). 3. adomgladathar 26’; cf. 30, Izimaoalladar,2 Eg. 30). labrnr. aralahdar, 28. sg. 3. Deponent preterite diuchastair ‘ awoke,’ 3.

Fled Bricrend 7 Loinges Nac n-Duil h-Dermait.3 aditgor. pres. ind. pl. 1. nitaghamar ‘we fear thee not,’ 199.

CniriIW. pres. ind. sg. 3. doacurethar, 126. pl. 3. doacuiretar 6edcl 53. pret. sg. 3. cotochrmtar ‘ so that it alighted,’ 259. fetar. sg. 1. n@tar, 186. 2. infetarsu. 156; cf. 135, infetarais,( 203. 8. fut. sg. 1 roe8sur, 157. 8. subj. sg. 2. cofesar, 106. clagiaam. pres. ind. sg. 1. adgladuraa, 231. 3. atagladad[ar], 217. athgZedar,s 198. pret. sg. 2. altrogailser: 231. 3. atgladastar, 110.

1 Eg. haa atzlmgladcstar, which is to be corrected with Windisch to atumgla- dathar ; cf. atomgladathar in the following line. 2 Lc. is here corm t. Edited by Win&ch from the Yellow Book of Lecan, Irische T6.q ii. 1, 164aq. 4 A Mid. Ir. new formation forfetar. 6 Aa to a possible adgladar, see p. 52, note. It is impossible to reprd a here as anythin but a blunder, due perhaps to association in the writer’s mind with glaidim ‘ fshout.’ 6 On the pet. stem ad-ro-gZiis- cf. Thurneysen, KZ. xxviii. 151 sq.; xxxi. 99 sq. THE DEPONENT VERB 1" IRISH-J. STRACHAN. 499

DEPONENTFORMS. ACTIVEFORMS. samlur. ipv. sg. 2. eamailte lat, 87. eamailtaiu (sic), 80. sieeiur. pres. ind. sg. 3. araiaethar (MS. -etar), 48. pl. 3. arsisetar, 17. subj. pl. 1. ardanesamar ' we will support them,' 44. Isolated form condarlaithir ' dass du sie triffst,' 204. Deponent pret. Tomarastar a ' remained,' 272.

Togail Bruidne Da Derga (LU. 83L998). 4gur.3 pres. ind. sg. 3. atcigethr, 87b 24. *b&iur. pres. hd. pl. 3. bid, 95h 30. *ciallnr. pres. ind. sg. 3. ciallathar (v.), 918 4. ClUiIliur? pres. ind. sg. 1. rocluiniur, 85" 15. OulriIU. pres. ind. sg. 3. docuhethar(bedg),87~30;\

14, i?&3W6&W, 84" 27.1 pl. 3. toacurstw toscur8- (MS. imda7acuiret,87b 20, tk),bedg, 86b 38. fetar. pres. ind. sg. 1. 921 ktur, 82" 27, roefeetar, 93, 11. fetw,O 85b 4, 928 26, etc.

Cf. tarlathar, p. 62. This verb in the Gilos8es and elsewhere is regularly active; de nent forms are found in the Laws, u. 316, 1. 18, mad marathar, 338, f? 18, muna marathar. 8 nisndigsimmis. of the facsimile 85b 25, should doubtless be nimiigjmmis. 4 Pres. 8ec. rocldtis, 858 13, fnt. sec. pass. rocechlastai, 88b 24. 5 iman-dath n-ecsamil docorethar fair; cf. 92. 26, Ldat ilg& docuirthir infolt fair. The pres. sec. is docorad, 85b 14. 8 This form offetur, with -ur from the present deponential ending, is common in these texts. So nifetwsa, Irish Notes in the Book of Armagh, 11. 600 THE DEPONENT VERB IN IRISH-J. STRAGRAN.

DEPONENTFOEME. ACTIVEFOEME. ipv. sg. 2. jintai, 92b 37. 8. subj. pl. 3. rofessatdr, 878 9. adglidur. pret. sg. 3. ataraglastar .i. roaicill, 868 17. labrr. pres. ind. sg. 3. intan kbras, 8gb 25. laimnr. pres. ind. sg. 1. ai taimim,l 9gb 9. 3. szilomethm-, 958 11. pl. 3, fdldimetdr, 958 10. samlur. ipv. sg. 2. samailts, 85b 4, 18, 86b aamail, 938 11, 948 5. 13, aamailti, 87b 21. eissiur. pres. ind. sg. 3. tairissedar, 898 3. tairisethar, 95b 23. “tuiginr ‘ cover.’2 pres. ind. sg. 3. imdatuigethar,z 9Sb 25. Deponent perfect, dochoemnacair, 988 28. Verbs in -ag-, -is-. Deponent, nodaiamaigetar (‘ place ’), 9Oa 27. Active, cressaigthi (‘ shakes it ’), 8g8 28, cresaaQth, 958 36, . rataiges (1. rathaiges ‘ marked ’), 83b 15.

Flight of EtLin (LU. 129). cuiriur. pret. sg. 3. condacorastdr, 1298 15.

1 In a version professedly taken from another source; it also contains rolm for earlier rolil. 2 Cf. ardatugethar ‘which coyers him,’ LU. 81’ 11 =LL. 78b 32 ; in a similar description, LL. 120a 47, rutuigedar ; in a quotation in Coma’s Glossary. S.V. Turigen, nitulach fri tuirigin tuigethar tuile mar mvirne, which may mean ‘ not a hill for a king whom a great flood of spears Covers ’ ; ef. Stokes, Trans. Phil. SOC. 1891-3, pt. i. 194. On analogy of the dep., rodligestk, 83a 12, a8 in M1. ; other dep. preterites rofersatar, 9’ib 42 ; im?mmda’tnestat.,9Sb 42. THE DEPONENT VERB IN IRISH-J. STRACHAN. 501

DEPONENTFOBMS. ACTIPEFOEkl5. muiniur. pres. hd. sg. 1. kmuiniur, 12gb 8. siasiur. pres. ind. sg. 3. assisedar (=ad+;), 1298 36.

Tochmarc EtAine (LU. 129b-130b). -tlnchur. pret. sg. 3. atlaigmstcir, 130b 17.' Deponent perf. nA wo%mair, 130" 5. verbs in -ag-, -is-, hdigth (pres. sg. 3), 130b 17, cairigius, 12Qb44, rhthzigis, 1308 13.

Tochmarc Etgine (LU. 12gb-132). Clnininr. pres. ind. sg. 3. chmdhm (r.), 132. 8. cniriur. subj. sg. 2. cuire, 131. 7. took8 (r.), 132. 7. ipv. sg. 2. cuir.&'(r.), 132. 7. Vo-ciallur. pret. sg. 3. foohiallastar .i. rotindil, 132. 16. Deponent perf. dci'mair,* 132. 1. Deponent pret. alra+estar, 131. 15. *gugSiniur. fut. sg. 3. dogignestdr do menma, 131. 19.

Trlin Bb Cirailnge (LU. 55-82, LL. 53b-104b). Bgur, do-dgur. ires. ind. sg. 1. atdgur, LU. 758 30.

In other LU. texts not given apart here tothZui,qestar, 128b 7 (Compert Conculaind), atlugestar, 133' 18 (ScCI Mongain) ; addawair, 1338 19, atu'dmuir, 133b 2. An isolated dep. pret. eissistir .i. iccrfaigis 'asked,' 134b 10. a There is B marked contrast bet,ween the LU. and the LL. texb in their treatment of verbs originally deponent. In the LU. version the deponent inflexion is well preserved ; in the LL. version deponent forms are hardly found Phil. Tran~.1891-2-3. 33 502 THE DEPONENT VERB IN IRISH-J. STRACHAN,

DEPONENTFORMS. ACTIVEFORMS. sg. 3. atotdgdhar (v.), LU. 57b5. ardattirgadar (v.), LL. 57b 29. nagathar who does not fear’ (v.), LL. 958 34. nitt&gadar,‘ 73h 9. pl. I. Xafpopar, fi~~~pt11Hal n&%@ar. In the corresponding passage, LU. 70%43 has the unintelligible nachitatss nachitchata. Considering the Irish fondness for alliteration one might emend nachitchaiw mwhatchra ‘who hates thee not, who loves thee not’ ; there might be a verb *cassim conueded with ~aiss‘hatred.’ Eg. 93 (Rev. Celt. xiv. 258) has the apparently corrupt nuchatarathar nacha- tawtthear. THE DEPONENT VERB IN IRISH-J. STRACHAN. 503

DEPONENT FORMS. ACTIVEFORXS.

subj. pl. 1. clunernni, LL. 5. 3. ncicocldrat,‘ LL. 95&18. ipv. sg. 2. clzdinte (r.), LU. 66b 38. concechZafat,2 LL. 618 fut. PI. 3. . t 41; cf. 64842. cobraim ‘1 help.’ pres. ind. sg. 1. nachatchobrathar (v.) ‘it helps thee not,’LL. 848 42. pret. sg. 3. cobrastar, LL. 87b 34. cuiriur. pres. ind. sg. 1. friscuriur,’ LU. 57b 26. 3. doczlrethar lights upon,’ gcuirmd, LL. 78b 33. LU. 71b 10. tocurethav (with ellipsis of bedg) springs,’ LU. 63b 23. focuirethar,LL.59,1.47. foch~ridar,~LL. 73b 49. PI. 3. curit, LL. 968 25. subj. sg. 2. cube( as ipv.),LU .5 88 17. cocorustur, LU. 5gb 32.4 diczcris, LL. 62b 27. pret. sg. 3. dochorastdr, Lu. 658 41 ; cf. 82&13. tocorastdr, LU. 708 10. ratchuz‘r, LL. 10!P 15. fut. sg. 3. icurfe (v.), LL. 56&48. immacurfi, LU. 72b 4. fetar. sg. 1. rafetursa,LU.71* 1,etc. and LL.

1 Cf. on the form p 6, note. 2 With transition to the b future. A simple reduplicated future with act. ioflex. noscechla is found, LL. 47“ 12, in a poem ascribed to Dallan If. More, the corresponding pass. crchluitir, 47“ 11. Pass., 3cechlastar, lOOa 25. a fochuridur alinta don chlettin for& sruth ‘the bronze of the spear lighted (?) on the stream ’ ; cf. LU. 70b 10, airm inarrasar anumu don cletiniu ‘the place in which the bronze of the spear rested.’ 4 Does corodechrastdr belong here, LU. 80a 14 (the passage is quoted, Windisch, To&. 66Ba,top of page), where the meaning is not altogether clear ? Mr. Stokes suggeste that it means ‘mered,’ comparing &chop. ‘difference.’ 504 THE DEPONEKT VERB IN IRISH-J. STRACHAN.

DEPONENTFORMS. ACTIVEFORMS. rafetar, LU. 6tJ8 33, etc. and LL. 54b 5, etc. 2. rafetarsu, LU. 6gb 30; cf. LL. 62b 44, 65b 6. 3. niconfitir, LU. 73" 12, eto. and LL. ipv. sg. 2.$ntassu, U. 55b 26. fut. sg. 1. jnnubsa, LL. 100b 50. 8. subj. sg. 1. co fessnr, LL. 6ae 33. pl. 1. co fassamnur, LU. 58" 18, 71" 4. *fochiallur. pres. ind. sg. 3. fdchiallathar (r.), LU. 78" 30. *follnur. pres. ind. sg. 3. fulnathar (r.), LU. 67b 4. *gainiur. perf. sg. 3. rogen&ir, LL. 85b 18. fut. sg. 3. adgignethar,'LU. 68'2. gigne, LL. 87b 35. adglitdur. ipv. sg. 2. acallsu, LL. 708 30. raacallais, LL. 70" 48, pret. sg. 2. 87b 29. argl&dais, 78" 15 3. ninarlastair,2 LU. ?la11. LU. =LL. 76a40. urZastur(rel.), LU.7 i840. laimiur. pres. ind. sg. 3. ni laimethw, LU. 67" 33. PI. 3. nalamat, LL. 72b 21. perf. sg. 3. ni ldmair, LU. 81& 41, 82" 33. fut. sg. 3. nocoltkaither, LU. 63" 15.

1 adyignethar dd dch ni atbath 'everything that was destroyed shall be made good to him'; ef. adgeznithir (sic) 'they are returned,' Laws, ii. 312, 1. 24, aithgin 'restitution,' ibid. 272, 276, 310, etc. 2 Cf. p- 55, note : arlastar under the influence of the dep. ending -astar ; cf. tawastaw, p. 29. THE DEPONENT VERB IN IRISH-J. STRACHAN. 505

DEPONENTFORMS. ACTIVEFORMS. midiur. pree. ind. sg. 3. midedar (r.), LU. 66b 21. perf. sg. 3. colzkwmaduir ('hit '),LU. 73b 20. damidair, LL. 80' 31, 102b 24. rfddair, LL. 92' 49. molnr. pret. sg. 2. rO8mOhi8, LL. 83'24 ; cf. 1.37. maramolais, LL. 83&22. sissinr. pres. ind. sg. 3. tarrasaaid, LL. 103&31. pl. 3. araiaetar, LU. 59& 29. subj. sg. 3. (as ipv.), faeiaedar he she11 stop,' LU. 718 34. ipv. sg. 2. airisin, LU. 32, LL. 76b 38. ddthraccar. sg. 1. dodhthacar, LU. 67' 29. pl. 1. dutliracmar, LL. 82&9. *tuigiur. mdatngethar,' LU. 81a 11, LL. 78b 33. Isolated deponent forms-faichlither, congebethar,2 LU. 7gaS 18, 81' 30, LL. 77' 3, 19 ; ~onidtharlathar,~'80 that it Cf. p. 67, note. * Taichkthar cath (?) (r.), 96a 37 ; cf. ndfaichledar, Saltair na Rann, 8066, and foichliwa, Windiach, Wb. 3 This form seems to he used impersonally (cf. p. 54), followed by do. LU. 7ga 18 (=LL. 77" 18), jbcheird a lhrecha . . . . . itnmd echaib congebethar doib o thul co awdornd do gdinib, etc., which ieem~to amount to ' he throws mailcoats about the horses so that thev are covered from forehead to crou with little spears,' etc. Eg. 93 (Rev. Celt. xv. 77) has lan doghainibh. Ld: 81a 29, cliabinar erdil congebethar dd co barrirachtar a dondfziathroci don dkrgi mileta do urol tig 'a cliabinar (breast-tunic) of silk, 80 that he is covered to the upper extremity of his fuathbroe with the warrior scarlet of the silk of ,kings.' The literal translation may perhaps be ' there extends to them (him) of. LL. 79a 4 has condriced. Cf. thara iathroig srebhnaidhi srdill sechtair congebhkthir dhdson o thana a thaibh go tiqh a shliasda, Eg. 93 (Rev. Celt. xv. 77) : cf. LL. 778 41. The third passage, magebethar ar lzimnairecht lcim dd anechtair, LU. 79s 3=LL. 77a 3 =conachgebedh ar lzimnnibh a lam aniaigh anechtair, Eg., is not clear to me. Active congeib ' which extends,' LL. 99%37. A transitive aidathar is found, LU. 71a 26, tdet Lugaid chuci 7 aranairlathar dd inn ingin, which seems to mean 'let L. go to him and offer (T) him the maiden,' but I have no other examples of such a meauing. 506 THE DEPOXENT VERB 1N IRISH-J. STRACHANL

(the spear) lighted,' LL. 73b 45 (cf. cotarla, 74" 37) ; conlind- fadar,' LL. 102" 48, 103" 21, imm~rsinithar~;of lodonethar, slzgethar, nerethnr (r.), LU. 78" 27, 28, I can make nothing. Deponent perfects-LU. conac&nnacair, 7 7" 33 ; conarrusair, 81b 26; inarrasar, 70b 10; tarrasair, 80* 39. A further extension by analogy, imchomarcair ' he asked,' 62R 7 ; ma- logical extensions in LL.? atacaemnacair, 97a 35; ritacoemnacair, 97&22; conarnecar, 80" 21 ; ni fmlangair, 87" 15 ; ni arlacair, 57" 2, 69" 28 ; forromair, 83&8 ; bnp&dar,68b 21. Verbs in -ag-, -ig-, Deponent LU. niairigestcir, 6gb 34; arigsetar, 57" 17; cotne'icmgedar, 61b 20 ; nonsarazjledmr, 59" 26 ; LL. radehyetar, 57" 16, rachorn&~gestar,~ 86b 7 ; rafeochaigsetar, 86" 6 ; goroguasaiysetar, 86b 45 ; rarzcamnaigsetar, 103b 54.5 ' From the contest the word seems to mean ' dare,' cia colindfadar inn dchain mo plhoba Chonchobair do thuarqain amlaidseo 7 m&si imbethaid who shall dare to strike my father Conchobar's "Groaner" in this way, and I alive?' ; cia colanfadhar na balcbriathraso dordda frimsa 'who shall dare to speak these mighty words to me ? ' But that is a mere guess, and 1 have no other instance of the word. * Impersonally immasinithar dhib ' they grapple with one another,' LU. 60%3 ; cf. p. 54, note. Atacrenanacair. ritacoemnacair are new formations for the Old and Middle Irish atachomnaic, which ia also found in the TGn. Conarnecar ris 'who met him'; cf. conarnic, 75b 17. N&rlangair 'he did not endure,' from fulangim. Ti adacair he did not permit ' from arleieint ; cf. ni arlaic, LL. 73b 29. Forromair ' he laid' =foy"fln dfuirm*?. As to imppa'dar-impadar Cuchulaind friu 7 benais a sL cinnu dzb ' C. turne upon them and cut 08 their sir heads-I can only suggest that it may be an analogical formation from the 3 pl. impdtar (LL. 103a 28). a Deponent forms are very common in LL. in the third persons of. the s preterite, and these map be retransformations of active forms. 5 Iu LL. de onential forms are exceedingly common in the s pret., especially after p. 80, an%above all in the Eerdiad episode. In LU. they are most frequent in p. 79, a passage which is found in much the Bame words in LL. The following list is not exhaustive. LU. rocarastar, 72a 31 ; eondawmnastar, 64a 9 ; dordtlaistir, 68b 15 ; rogabaastar, 79 passim ; roinnisimar, 588 4. LL. &lastar, 63b 25 (radd, ,64a 11) ; rairastar, 81a 33 ; nirfaemastar, 74b 39 ; raeittchestar, 818 33; ragabastar, 72b 48, 77 passim, 86 passim, 92b 4; rogdsistar, 618 18 ; nirragonastar, 1038 51 ; rognuisestar, 618 18 ; rarmar- nestar, 85b 19 ; romahastar, 72b 48 ; ratmelestar, 86b 21 ; ratnascestar, 86b 23 ; rudestar, 74" 32 ; rarepestm, 81. 33 ; rosniastar, 58a 40 ; rothigestar, 86b 20 ; ctarasredestar, 68b 47 ; ralregdastar, 86b 22 ; tucastar, 7% 34, 1038 28 ; rouvg- nastar, 638 23 ; nardermamar, l00b 6 ; ratasairgsemar, 100b 6 ; rahertsatar, 91b 28 ; bachoml~risetar, 95% 10 ; rabulgsetar, 104a 2 gorachloesetar, 84b 4 ; conrochrothsatar, 98" 7 ;radumsatar, SOa 40 ;gorachmrarcsetar, 86b 39 ; rachrr m- satar, 80b 32 ; rafucsatar, 96. 51 ; faitsetar, 91%26 ; gorojillsetar, 86b 45 ; focherdsetar, 84b 11, 35, cf. 85b 40 ; gabsatar, 68b 50; ragahatar, 55b 10, POL 25, 84a 50, 84b 17, 85%8, 8329 ; goragdrsetar, 86b 48, cf. 82b 46 ; bogni- sstar, 84b 32 ; daronsatar, 86b 39, 42, 44, 47, 51, 87' 5 ; dariripetar, 87" 3 ; gorolasctar, 86b 52 ; condaralasatar, (MS.-astar), 73b 32 ; gorol&psatar, 86b 45 ; raraidsetur, 79b 28, 808 48, 91b 3, 92a 20; atraigsetar, 89b 9; rai&%dsetar, 8Ob 30 ; coroscaigsetar, 92b 22 ; rascoirsetar, 84b 11, 27, 85a 26, 85b 39 ; ratheigsetar, 91a 43 ; rucsatar, 898 4. THE DEPONENT VERB IN IRISH-J. STRACHAN. 507

Active LU. airigthi, 57b 2; arigit, 62' 23; bertaigthus, 61" 32; bertaigthi, 61" 36; bertnaigis, 7Tb 32 ; chaignait, 5gb 34; rachasnig, 80" 7; cresdigis, 7Tb 32; orithaaigset, 7gb 24; dddigmit, 5gb 34 ; LL. arigis, 86b 18 ; niairgem,

61b 47 ; rambertaig (6 he shook himself '), 54b 29; &artaigis, 6gb 48 ; bertnaigis, 64b 27, 76&10 ; buadaiggfes, 82a 38 ; racanig (1. -casnig), 78" 4; certaigis, 64b 29; coraigis, 66b 3; cdraig (ipv.), 668 49; cressaigis, 76& 9; crithnaigset, fTb 25 ; czctlaiga'e, 6gb 48 ; rachrkchtnaig, 84b 24 ; fiacumnig (ipv.), 84' 41 ; dehgjEt, 57' 17 ; roscomdluthaig, 62b 13 ; radorchaig, 85" 51 ; rdadbaig, 88" 22; fkdbaig (ipv.), 88" 20; Izir$ulig, 72" 29, cf. 84b 24, 80b 27; raminaig, 96b 43; vathaigid, 63b 45 ; ralhaigis, 5?b 23 ; rats.draig, 102b 18 ; ni sdridgum, 5gb 38 ; rasuanm$, 85' 52 ; ni thmnanaigfit, 90b 43.

Mesca mad' (LO. 19*-20b, LL. 26Ib-268'). DEPONENTFORMS. ACTIVEFORMS. cnirinr. pres. ind. pl. 3. na wire$, LL. 265' 6. pret. sg. 3. rmhuir, LL. 265b 25, fetar. Sg. I. nadfetar, LL. 264" 25. 2. infetarsn, LL. 264" 24. ipv. eg. 2. finta, LL. 263b 5. 8. fut. sg. 3. mjestar, LU. 19" tI. againiur. perf. sg. 3. rogenair, LL. 264. 46. adglbdur. pret. pl. 3. curamdket, LL. 26ib31 fut. sg. 1. atagegallarsa. LU. lgb30. 3. atagegalldathar, Igb 30. Qtageglathar, lgh 33.2

1 Unfortunately the end of the tale is not preserved in LL., while in LU. only a fragment of the end has been preserved. The two versions seem to stand in much the same relation to one another as the two texts of the rain, and they have been treated in the same way. A noteworthy feature in the LL. text is the frequency of -end so-called consuetudinal forms. 2 adgeglathar would stand in the same relation to *a&egladathar aa adghdar to adgladadfir, p. 62, note. 508 THE DEPONENT VERB IN IRISH-J. STRACHAN.

DEPONENTFORMS. Amw FORUS. labrur. pres. ind. sg. 3. ni labair, LL. 268" 5 (bis). laiminr.' pret. sg. 3. dr narlm, LL. 263" 5. sissiur. pres. ind. pl. 3. thriait, LL. 265" 10. Deponent preterites *-LL. ferastar, 262" 28 ; rdirastar, 262b 1 ; ratalineetar, 263" 8 ; rainnisetar, 268b 9 ; tincsetar, 262b 21. Verbs in -q-,-is-. Active forms-LU. stcdigthi, 19.21 ; LL. condaabethaagend, 266b 35 ; na hkligaiu, 265" 19.

Serglige Conculaind (LU. 43-50). 6-r. pres. ind. sg. 1. ~i agar, 44" 35. pl. 1. atagamw, 48b 27. subj. sg. 2. ni aigthev, 44" 36. nitdgara(r.), 49. 33. Wrliur. subj. sg. 2. airlb?her,4 46b 9. cnirinr. pret. sg. 3. rochuir, 47b 14. fetar. subj. pl. 3. jnnatar, 46b 11. s. eubj. sg. 1. cofiaaur, 45" 26. 3. mhfeatw (v.), 46b 32.

1 ni kmthar, LL. 2668 27, is passive in form, and may be translated aa a paasive ' it will not be dared.' 2 Instead of the actipe tarnaic=roscdich 'waa past,' e.9. LL. 57' 13, trdth tarnaic do chdch wrgnom bid 7 lenna ' when the rest had finished preparing food and drink,' we find in this text, 262b 49, nithamacar acht a ewh do ocur 'his horses had only just been unyoked,' 262b 31, coturnmar Zeo. 3 Ehted by Windisch, Woche Texte, i. 206 sq., with vsrianta from H. 4. 22, T.C.U. The iece contains the Briatharthwc Cmulaind, LU. 46b 1-30. 4 So I wout)d read for airliter of MS. : airlither mmni 'thou shalt consult memory or tradition.' In the same passage the obscure mrogatar seema to be deponent, mogatargendaigi gesci &a geinither (soH. 6.22, wrong1 geniter, LU.) gan. MI. Stokes calls my attention to broga .i. $renuigther, &Don. Snppl., and suggests that the meaning may be 'let them verify the branch of genealogy,' etc. THE DEPONENT VERB IN IRISH-J. STRACHAN. 509

DEPONENTFORMS. ACTIW FOBMS. *gainiur. pres. ind. sg. 3. genitirer,' 46b 12. adgl0dur. pret. sg. 3. acallak, 45b 36. labrur. pret. sg. 3. labrais, 44b 14. laimiur. pres. ind. sg. 3. ldmthair (r.), 49' 38. -ciu. (P) subj. sg. 2. faitchither,* 46b 17. Yerba in -q-,-$-. Deponent s-wigsitw, 448 17. Active-roakrdgsst, 49' 24 ; crhhtnaigid (r.), 45b 5 ; wht- chuibdig (ipv. sg. 2), 46b 30 ; ragnathaigeem, 45' 4 ; rordtha$, 49' 25, 4gb 43 ; dianaithiget (v.), 47b 18 ; taithiges, 47" 14 ; romthathigtet, 47a 2.

Fled Bricrend (LU.9gb-1 12b4). ad-0gur. pres. ind. sg. 2. ataigthr.6 cluininr. pres. ind. sg. 2. atclunisiu, 104b 9. CUiriIU. pres. ind. sg. 3. curethw (r.), 106' 41 ; nichuir (r.), 104b 12 ; cf. 106b 33. cf. 108b 18. subj. sg. 2. mire samla, 10Fjb 30, 106' 18. ipv. sg. 2. ouir a samail, 106b 3.

1 LU. gmiter. See preceding note 2 nifartchzther 8morr. O'Currg translates ' You will not deride old people.' If the form be correct it might be a subjunctive from a compound, fo-aith-chiu ' look down upon,' ' despise,' but this is a mere guess. Deci, &a, as ipv. 47b 2, 3, 4, etc. (v ). On analogy of deponent rocharastar, 4Sb 12, 13, 15 ; CWOnkaiChneStar (0. Ir. arthged.wa), 488 5 ; roscarsatar (v.), 47% 33. .I Edited by Windiseh, Irische Tezte, i. 234 gq., with variants from Egerton 93, Brit. Mus. and H. 3. 17, T.C.D. The concluding portion, Cmnach ind Ruanada, which is incomplete in LU., has been pubIiRhed from a Leyden MS. (L.) by Stern, Rev. Celt. xiii. 28sq., and from an Edinburgh MS. (Ed.) by Meyer, RC. xiv. L. =attaidiisi, Ed., Rev. CeIt. xiv. 454 ; 80 adaghatctir, L., RC. xiii. 31 E. attadar, Ed., RC. xiv. 453. 510 THE DEPONENT VERB XN IRISH-J. STRACHAN.

DEPONENTFORMS. ACTIVEFORMS. %omalnur. pres. ind. sg. 3. comaZlas, 1 lob 12. esur. 8. subj. ag. 1. condaeaur, 104* 14. fetar. subj. pl. 1. jinaamcir, llQb38. adglidur. pret. sg. 3. atgZridmtm, 9gb 35, 1Olb 34. laimiur. pres. ind. sg. 3. Zairnethur (r.), 102b 15. perf. sg. 3. rotrimair, 112' 37. "folamur. subj. sg. 3. folimathm, 104a38,1 110& 49. pret. sg. 3. folmaatar, 102a 20. pl. 3. folmaiset, 103b 31. midiur. * pres. ind. sg. 3. dmmidethar2 'he tries,' lllb 45. PI. 1. cotmidem, 110' 23. 8. fut. sg. 3. conmestlur3 (r.), 104a 7. moiniur. pres. ind. sg. 3. forathmmadw, lllb33. fwaithmenathur (NS.-atar), lllb17. perf. pl. 3. dwumdmatrir, 11 la 34. molur.

I pret. pl. 3. molaat, 10Bb 40. diithraccar. Bg. 1. donkthrtacm, loob 10. tuigiur. pres. ind. sg. 3. fwdatu&$har,' 105b 44.

1 E has here rotlemathar,fut. of ro-Zaimiur. 2 Cfp. 15, note. For miastar the fut. pass. cotmiqder, 112b 23. 3 ni faigbistar fer and conmestar a as 7 a &, etc., 'there will not be found a man there who will equal his age and his growth,' etc. Here faigbistar seems, as Mr. Stokes also has suggested, a barbaric u future from fogabaim. 4 infolt fordatuigithar 'the hair which covers him'; cf. p. 57, note, and add fordotkigithur, Laws, ii. 284, and otmtuigithear, 11. Texta, iii. 18. THE UEPONEST VERB IN IRISH-J. STRACHAN. 511

. itdcin. subj. sg. 3. ninaccalhar, 40. Deponent perfects-furcdemnacair, I 11'" 39 ; ardamair, 112" 41 ; ni furddmair,' 111" 12. Isolated deponent forms-coihlethar (r.), 1 02" 7 ; dligetar (r.), 109" 35; fwdsnadar4 (r.), 106* 44; gebitar5 (r.). 104" . 41 ; nosintthar,6 110" 31 ; immacomsinitar,' 105& 2, 109" 16 j immastecrathar (r.), 102" 1. Verbs in -ag-. Deponent forms-rofergaigestm, I Olb 3 ; inrabrefhaigestar, 112" 40.9 Active forms--namzi hrelhaigeseo, 107b 28 ; brethaigfetsa! 107b29; chat[h]aigmitni, 107" 22; rocrithnaigset, 101" 2 ; two- dilsig, 10P 30 ; fortamlaigid, 109" 5 ; Izirrathaigaem, 10f1~11.

. Eiaburcharpat Conculaind lo (LU.11 3a-1 1 fjb). clnininr. pres. ind. sg. 3. rochluiIzethar (v.), 115" 3. subj. sg. 3. coclothar (v.), 114b 43.

1 nirdaimaet for nivdamatar, 110b 35. * The precise meaning of the word ia uncertain. The general sense of coi62ethav ce'im seems to be 'who takes a leap'; cf. coibliud bhada, LU. 10P 21, and dr atchota ilchumachtai ocai fri coibled fergnim, LU. 123b 1. Qaibliud may stand for com-$Zlizul, and me might perhaps compare filliud evred ndir, one of Cuchulinn's feats. 3 Read with Eg. is Cuculainddligethar 'it is Cuchulinnthat has a claim to it'; for the deponent form cf. dligkdir, LL. 396b 30. 4 Windisch seems right in taking fudsnadar here and diafuasnaither, 106b 2, as deponent forms used in a middle ~enae. The passive form in the Glosses is fuasnither, amaE fufuasnithm, w1. 66d 17. 6 Read ge6ithar with Eg. From natfri goi gebithar of Eg. we should robably emend LU. to nod fri gde gebithar ' who is not given to (P) falsehood.' bhether that reproduces the sense or not, gebithar seems to be a deponent form. 6 In a middle senae 'he stretches himself '; the same meaning is expressed by the affixation of a pronoun to the active form in sinithi, LU. 59b 39. 7 This should be inamacomsinithar ddih. For the construction, cf. p. 54, note. 8 immuslecrathar 'covers them,' ' protects them.' Cf. MI. 65%1, iarsi7tdl adcuaiduom dineuch imme thecrnthar crist dianechtair otoi talmaidiu duaisndis de fessin hic 'after he has spoken of all that cover8 Christ ythout (i.e. his outward appearance he suddenly turns to speak of Himself. Cf. the dep. inflexion in the wori *tuiggiur, of kindred meaning. ' Other deponent preterites are roderseaigestar, 99b 9 ; iarlastdr (Zdaim 'throw'?), 102a 28; itallastar, 107" 30 ; cf. llla1. 28; fait6e8tdr7 100b 13; ochsatar (r.), 102b 24. lo Zimmer, Zeitschr. f. deuteches Alterth. xxxv. 43, dates the text from the middle of the tenth century, but it is probably earlier. Kote the preservation of the old subjunctive stern clo-. $12 THE DEPONENT VERB IN IRISH-J. STRACHAN;

DEPONENTFORMS. ACTIVEPOEMS. *gainiur. perf. sg. 3. rogenair, 115' 37. adg1tidur.l subj. sg. 1. conidnarladw, 113" 7. pret. sg. 3. ni adasair, 114&3. Verbs in -q-. rochruthaigeetar' (v.), 115" 16.

Echtra Condla (LU. 120n-1219. cuiriur . pres. ind. sg. 3. totchurethm (r.), 120" 43. pret. sg. 3. dochmastdr, 120' 34.

aagitiaw.- whom thou ad- pres. ind. sg. 2. dressest,' 120' 13. 3. adgh%d@r,120e 14. Verbs in -ig-. cotomdimigidar ' compels me,' 120' 28.

Aided Conculaind (LL. 1 19"-123b). ad-hgw. pret. pl. 1. adra&~senw,120' 29. "airliur. pret. sg. 3. cmairle8tar, 123b 3.5 "buriur. pres. ind. pl. 3. comhwetar (r.), 119" 19. "ciallur. pret. sg. 3. c~dla~tar,~123' 31,

1 In 115. 23 atomglaithe would formally be most naturally taken for an imperative=atonzgZddithe, cf. p 52, note (but atomglaite, p. 48), but from the meaning it is questionable whether the writer did not intend it for an indicative. Loegaire refused to believe St Patrick till he should have seen and spoken with Cuchulinn. Cuchulinn now says to him, utomchi a Loegairi utomglaithe Zetr ; mani eretea Putraic biasu hi pdin thou seest me, Loegaire, thou speakest with me face to face; if thou believe not Patrick, thou shalt be in pain.' Here an indicative seems much more likely than an imperative. Possibly we should correct to atomgzaither. A deponent preterite beogastar (v.), 115* 14. 3 All in the lament of Emer over Cuchulinn. THE DEPONENT VERB IN IRISH-J. STRACHAN. 513

DEPONENTFORMS. BCTIVB FOB,^. *cobrr.l nachitchobradar 'which ind. sg. 3 (r.) does not help thee,' 1 119b 20. ipv. sg. 2. cobairthe (r.), 119" 17. pret. sg. 3. cobrmtar,' 123b 14. conoim. pres. ind. sg. 3. comcathw (r.) which protects ' (?), 119" 29. pret. sg. 3. formastar,l 12P 7. CniriUr. pres. ind. sg. 3. domridar bedg, 11gb 5, 121b 16; without bedg, 1228 2; cf. 4. *follnur.S pret. sg. 3. fallnastar,' 123b 6. *gainillr. fut. sg. 3. gignithwr (r.), 119b 48. ad-glbdnr. pret. sg. 3. acillestw,' 123b 6. laimiur. pret. pl. 3. ni rolamsatar,' 121b 22. sissinr. fut. pl. 1. aurksfemmar, 120b 13. -tluchur. pres. ind. sg. 3. do.~othhg,122" 45. *tniginr. pres. ind. sg. 3. ratuigedar, 120a 47.

1 Other instances of deponent inflexion in thL, verb, in addition to those already given, are nimchobrathar 'it helps me not, LL. 284b 26 in a story of St. Molling), nomchobrudarsa, 286a 28 (in a story of a sister of NoI asse). 2 Deponential forms from this verb haye been given above, p. 51 ; as a rule it is active. 3 fdlnabthair, 123' 1, may possibly be deponent, but the passage is not clear to me. 4 Remodelled after the B pret. from the old rolamatar. 511 THE DEPONENT VERB IN IRISH+. STRACHAN.

Deponent pe~ects-inichoemnacair,' 25 ; nadfoordamarsa,2 119&1 ; tarrasair, 121b 6. Verbs in -ag-, -is-. Deponent forms-roSdraigestm-, 122b 32 ; miada@estar,8 123h 9 ; rathaigestar,3 lBh11 ; urdaircaigestar,3 123h 10. Active forms--crechtn,a@$t (r.), 1 1g8 40 ;rachraig, 121b 35 ; fadhiget (r.), 123&5; raroSCiraig, l/gb 16.

Tochmare Emere (LU.121e-127b). DEPONENTFORMS. ACTIVEFORMS. Wuiniur. pres. ind. sg. 3. atchluin, 126" 18. *comahur. pret. sg. 3. comallastlur (Facs. oom- allastas), 127a 17. adgltidnr. pres. ind. sg. 1. adglddw (Facs. adgadur), 124" 4. molur. pres. ind. sg. 1. co ro'omolaim, 124" 23, auidiur. pres. ind. sg. 1. fos~uidiur,~1248c12.

1 In this passage imohoemnacair gives no sense. Perhaps we should read imchomamair ' asked,' which would 8Uit the context. 2 From dmim ' grant,' 'give ' seem to come the new analogical formations Birodamsat, 122b 30 ; rodamastar, 122b 31. 3 From the lament of Emer, in which is also found the deponent preterites briigestar, 123b 15. 4 The word is found in the same phrase, Windisch, Ir. Texte, i. 141, 1. 26, fosuidiur a n-dano 7 a n-dibergai. and in the Laws, ii. 380, 1. 12, fosuidither ; cf. 1. 13,386, 1. 6; 16, 18, 19. The deponential inflexion still prevails in the text of tlie Laws, a fact which must be taken into account in'fixing their date. I have noted the followin forms in a cursory examination of the text; whether the blunder8 are iue to the copyists or the editors, the method of printing makes it impossible to tell without an examination,of the MSS. (the printed texts are very untrustworthy)-imnzuscobrathar, ii. 280 ; nodacmallathar, iii. 32 (but comaZlat, ii. 140, 306) ; inidcuirithar, incuirither, iv. 168, cf. ii. 284, 290, 306,. 323 ; rofallnnstar, iii. 30 ; arafeiser 'that thou mayst know,' iii. 106, 126, etc. ; co$nnathar. iv. 190, cf. iii. 8, 60; genither, iv. 38, lasanaithgenitm, iv. 164, cf. ii. 312 ; labraither, ii. 306, etc. ; each lazmethar, iv. 192 (corrupt rochlamethar, iv. 190) ; marathar, ii. 316, 338, 394; con- mididhar, iv. 16, etc. ; domidider, :. 320, cf. 142, 232, etc.; taivisedtnr (1. tairisedar), ii. 324 (but mani tairiset. ii. 328) ; dofeisedar (=do-fa-sissedar ?), iv. 62 ; dotluigigter (1. dotluigedw), U. 316 ; fortiotuigither, U. 284 ; airigtir (1. aivigetar), iii. 8 ; dilsi.qthir (1. dilsigithir), iv. 168 ; suidegetar, iv. 190, cf. 180, 188; toltanaiytef. (1. -eta?-), iv. 64; ronuaistigthur (1. -ethar), ii. 12. THE DEPONENT VERB IN IRISH-J. STRACHAN. 515

Deponent perfect (rather 8 aorkt)-twraaair, 126l 26. Verbs in -q-. nosbertaigend, 127’ 27 ; rockechtnaig, 126a 29.

5. SHORTSTORIES OF A RELIQIOUSCHARACTER, LL. 278 aqp. Digs macclerech, 27S8 37. DEPON~TFORXS. ACTIYEFORKS. clnininr. pres. ind. rotchlunim, 278b 19. sg. 1. 3.I nf chluinsthar, 278b 17. Verbs in -ag-, -Zg-. nachamchairig (ipv.), 278b 15 ; corothathbeoig, 278b 30.

Ri irissech, 278b 32. atlochur. pret. sg. 3. ni conroatla(gesiar, 279a 1.

BB ri amra, 2798 35. clnininr. pres subj. nactdrea, 2808 19. sg. 1. CUiriW. pres. ind. ) dochwedar, 279b 6. sg. 3. subj. sg. 2. cure, 279b 17. fetar. 8. fut.sg.3. rofestw, 219b 25. Deponent a preterites-roscarastr, 279b 5 ; ni roplend- 8emmar, 2808 10.

Dep. 8. subj. comairsw. iv. 18. Active, airtiut (if the redi be right), ii. 82; drmoigid, iii. 34; samaigas, iv. 8. On the other hanf the so-called consuetudinal ending -end is sometimes found urnaidend, iii. 54 ; ni obund, ii. 316 ; indarbann, ii. 306. l On p. 121b 32 is found nirodsathar arengn. The meaning must surely be ‘his beard had not grown,’ for it was a constant reproach to Cuchulinn that he waa a beardless boy (cf. LU. 74b 33). Of the word ren.ga in this sense I have no further example. If it be 8 singular, rodsathnr would be a form of the same kind as imp’dar, p. 63. But it is possible that renga is plural, and that rotisallbar ia an error for rociratar. 516 THE DEPONENT VERB IN IRISH-J. STRACHAN;

Cethrur Macclerech, 281a 40. DEPONENTFoaars. Amvs Foaus. cuiriur. PEE. ind. ocuirethar Qedg, 281” 13. sg. 3. -tluchur.

sg. 1. Deponent s preterite-cotairahellaat&, 2818 51.

Bdi ri amra, 2828 32. olnininr. pres. ind. ] ni chlukethw, 2828 46. sg. 3. -tluchur. at~ochontar,2826 40. pl. 1.

6. REGLUMMOCHU~A RATEIN,’ LBr. 261h262’.

1 Such is the title whieh’this piece bears in the Leabhar Breac. Mochuta of Raithin died A.D. 636, O’Curry, Manuscript Materials, 374. But in a manuscript in the Royal Irish Academy, 23, N. 10, p. 82 sq., the super- scription is Fothad nu canoine .cc. ham regulam. Fothad died, according to the Annals of Ulster, in the year 818. 23, N. 10 (N.) consists partly of vellum, partly of paper, which is rapidly disintegrating. The Regula stands. in the paper portion. The manuscript aa a whole has been copied from good sources, and in many cases in the Regula it shows a be’tter reading than LBr. Tbe duties of a king stand before those of a bishop, and are set forth at greater length than in LBr. At the end are five additional quatrains, beginning Btamba trebthach ba trebor ba fuarrach fri cwh, ba failidJ6 fri haidedu cia tisat gach trath, and ending Figell aim ernaigedi alrnsan tarr nosgene, nibar adbchlans dodoene ba ardia gach andene. Then, p. 88, 4, follow more precepts of the same kind, beginning Coma viaguil imoimded is ann ni faigae baoga2. The following points in the language may be n+ed :--deerc is still disapllabic, N. 84, 3 = dercc, B. 261a 42; N. 85, 25=&z’oepi, B. 261b 23. The n. pl. masc. of the art. is in, preserved in anbradhair, N. =na bratair, B. 261b 85 ; n. pl. neut. saa hibalru, 262a 26 (nu hilgaluir, I?.). Gen. sg. and nom. and acc. pl. fern. of art. are nu, gen. pl. more frequently .nu than inna. The termination of the dat. pl. of the adj. is preserved, lathib techtaidib; for cu frasa diana, B. 261a 35, N. has frosa dianuiph, whence may be deduced the reading frosaib diaBaib. Acc. of rigain is rignui, N. (=rigan, B. 2618 48). Comparatives sruthiu, siniu, superlatire asaeairem gle’s, B. 261b 76=isagaenem gles, N. (cf. gaoim ‘good,’O’R.). For danaine, 261b 75, N. has di aeine; for imi, 261a 29, N. has ani. The verbal system is well preserved-uaririsiu, 261s 23 (asriaszs, N.), THE DEPONENT VERB IN IRISH-J. STRACHAN. 517

DEPONENTFOEMS. cluiniur. prea. ind. sg. 3. ~.oscZuinithr,~261b 84. aubj. pl. 1. intan cloamar, N. 86, 1. 10 (&mar, B. 261b 38). comalnur. pres. ind. sg. 3. natcmdla,2 2618 44 (nadcomall-,N.). oomadlkus, 2618 39 (cmollt~,N.). cuiriur. pres. id.sg. 3. doohurathw: N. 87, 1. 8. subj. pl. 1. tochram: N. 86, 1. 10 (=tocham, B. 261b 40). adgladnr. pres. ind. adgbdamw, 261b 44. pl. 1. 1 atluchur. pres. ind. 2628 5. pl. 3. atla&t,b iaisroirr, 2618 66, dinlei, N. (=diaMg, B. 261b 9), ~pocomtuis(1. nococ~tii~), 6 thou shalt not be indulgent,’ N. 84, 22 ; B. 261L 65 is corrupt. Of 8 forms in 3 sg. s pret. only timarnus, 2618 29 (timamad, N.), but rotsoer, ronciin. For marbud tiugla, 261b 67, N. has amal tiughlaithe. Note also imalle, and inge ‘hut’ ; also the word arcofuin, N. (arczchjuin, B. 2628 l),explained in Cormac’s Glossary. In some cases later forms may have crept into both texts. Taking these things and the deponents into consideration, we shall hardly be wrong in putting the text at the latest in the beginning of the ninth century. In some parts it reminds one of the Regula ascribed to Maelruain, LBr. 9b sq., also an old text, though it chances to contain little deponent materid- dscuirither, 9b 10; cf. gb 12, 35, IOb 2, conothar (condim ‘preserve’), 10b 41. In LL. 148 we find a number of precepts professedly addressed by Fothad to Aed Oirdnide on his coronation, 793 A.D. But the language shows that the composition is of later date, na in nom. pl. masc. for ind, cluin ‘hear,’ nomwzolam, etc. Note also the reference to the Gail1 supposed to be in the service of Aed, 14Sb 21. 1 rachluineth-, N. But a plural is required-racluinetar. 2 261b 10 B. has a dep. comlltar=comallo, N. In B. the line is a syllable shokech natcomalltar so : N. has nech nalcomallo ansdn. 3 B. 261b 76 has less well docorustar. 4 Active, as already in the Glosses ; see p. 9. If this be right, the active inflexion may probably be ascribed to the inffuence of verbs in -aegim. N. has altaiget, which would be ipv. ‘let them offer a blessing.’ Phil. Trans. 1891-2-3. 34 518 THE DEPONENT VERB IN IRISH-J. STRACHAN. duthraecar. s fut. sg. 2. duthraisir (variant -er), 2618 9. Verbs in -ag-, -$-. Deponent-nifiGiZigthw? thou shalt not shed blood,’ 261* 7. Active-romiscneged, 26 1b 37 (better N. nonmiscnegat), athaigem, subj. (aithegem, N.), 261b 38. Deponent 8 preterite-rotordnsstar, 2628 15.

7. FBLJBEOF Omons. DEPONENTFORXS. ACTIVEFORXS. frisaili~r.~ ‘Iexpect,’ Ep. 560. clnniur. ipv. sg. 2. cluints, Ep. 314, 365, 425. fetar. s. subj. sg. 2. rqeiser, Feb. 4. dianfesser, Oct. 24. “gainiur. perf. sg. 3. roginair, Sep. 21 ; cf. Dec. 25. *labrur. nalllabrai, Dec. 22. laimiur. perf. sg. 3. rohmair, Prol. 58. molur. pres. ind. sg. 1. nomolur, Prol. 13. pl. 1. nosmolammar, Jan. 17. molmai (molma), Bug. 31. ipv. sg. 2. molatha, Sep. 2.

1 N. has less well dzcthmcar. 2 N. has ni chuilide. For czciEigde? cf. LL. 346b 38. 3 So Rawlinson ; Laud and Leabhar Breac havefrisailiu. 4 Aedan ingrian gelda inse medcoit molma, ‘Aidan, the bright sun of Inis Medeoit that we praise.’ Stokes, after O’Clery’s molma .i. molta, took inolma as a participle. But I know of no other instance of such a participle in Irish, and molmai may very well be 1 pl. absolute of molur. aS to O’Clery’s loss it probably arose from a misunderstanding of this passage. [I am glad to fnd ;hat Mr. Stokes agrees.] THE DEPONENT VEKR IN IRISH-J. STRACKAN. 519

-muiniur. pres. ind. sg. 2. admuinter,' Oct. 2.

An isolated dep. rimther ' thou countest,' is, according to Stokee, found Prol,. 286, Ep. 46. As to conruz'dz'ur, Prol. 277, it seems to be an instance of the dep. 1 subj., which we shall speak of later. It may be analysed into con-ru-suidz'ur. It is explained as 'may I attain to'; cf. Skr. pa-sad gut von statten gehen, gelingen. Deponent perfect damair, Feb. 9, cf. Feb. 16. Verbs in -ag-, -ig-:- Deponentma ncebraigther ' if thou remember,' Mar. 2.

8. TRIPARTITELIFE OF 8. PATFXCE? In this text, as in those that follow, if we leave aside for the present the 1 sg, subj., deponent forms are found for the most part in the perfect and the s preterite. The following verbs are still deponent :- agm. pres. ind. sg. 1. attdqw (v.), 114. 26. cluiniur. pres. ind. pl. 3. rochetar, 120. 13; subj. sg. 2, cocloithersu, 244. 12. fetar. sg. 2. infetarsa, 128. 7; 3, rojtz'r, 30, 25; pl. 1, rofstamar, 42. 11. laimnr. pres. ind. sg. 1. ni lomar, 166. 2.

1 So Mr. Stokes. The glossator takes it as 3 sg. pass. adamraigther m, Qendaich[tir]er'is admired' or 'is blessed.' As tojdiZi it can hardly be acc. sing., for that, in the PBlire, is regularly je'iZ ; see index to Stokes' edition. It might either be nom. pl., as Ep. 18 (rhymes with pl. c&), or acc. pl., as Ep. 108 (rhymes with eoZZiri). In tbe former caw fkZi would be subj. to 3 pl. pass. admuinter ; in the latter obj. to 2 sg. mbj. act. admuinter. PeZi could be taken as acc. sing. only on the supposition of a transition to the inflexion of i stems ; cf. Thurneysen, KZ. xxViii. 147, but this, as we have seen, is contrary to the usage of the F6lire elsewhere. 2 The references are to the pages and lines of Stokes' edition. 520 THE DEPONENT VERB IN IRISH-J. STRACHAN.

%folamur.l pres. ind. sg. 3. folamadair, 112. 8, 226. 10; folamadm, 208. 4; pet. sg. 3, folamustnr, 78. 11, 146. 20; fola- mastar, 80. 1, 156. 2, 168. 7, 200. 21; foh- mustair, 148. 28.

The following verbs have, except in the -ar perfect, assumed the active inflexion :- cuirim. ipr. sg. 2. cuir, 56. 4; pret. sg. 3, dora[th]chuir, 158. 20; fut. sg. 3, ni chuirfi, 78. 9 ; doaithkrfe, 158. 14. follnaim ‘ rule.’ a pret. pl. 3. rofallnaiset, 196. 20. genim. fut. sg. 3. gignid (v.), 150. 7 ; gigness (v.), 154. 18 ; gignes (v.), 216. 17; yenfess, 158. 13;3 pl. 3, yen& 58. 12. adgl4daim. pret. sg. 3. rosagaill, 114. 6. midim. fut. sg. 3. midfes, 260. 17. -moinim. pres. ind. pl. 3. noskrmanat they forget it,’ 82. 19. - sissim. pres. ind. sg. 3. tairisid, 8. 11 ; fut. pl. 3, tairisfet, 252. 27. -tluchim. pret. sg. 3. dwothlaig, 10. 184; PI. 3, dorothluigset, 102. 19. The deponent subj. of -ciu : conaccomar, 102. 12.

1 Stokes in his glossary translates this doubtfully by ‘ I desire,’ cf. lii. 1. 27. The verb is the same as foldmur, which we have already had in the sense of ‘swipio,’ ‘tento,’ pp. 13, n. 4, 67. On p. 112, 1. 8, of the Trip. Life, for folamadutr Patraic congabad cathair should be read folamudair Patraic congabail cathrach, similarly p. 156, 1. 2. 2 Another example of this verb is the sec. fnt. nudusfoilmaibed, 188. 22. In nocha%follamnatget,94. 26, follnaim is replaced by the derivatiTe follammatggim. 3 The prose shows the later genfes (cf. nogenqitis, 86. la), while the older gignes is preserved in verse. In the Lebar Brew text, Trip. Life, 478. 24, giggnid of 150. 7 has been replaced by gen$d. 4 The dep. dfwothlaigestar, 30. 6, must be taken in conjunction with the other deponent s preterites of this text, and need not be regarded as a survival of the old deponent inflexion of thie verb. THE DEPONENT VERB IN IRISH-J. STRACHAN. 521

Deponent peffects-forooimnacuir, 34. 16 ; ni cWmnacair, 126. 9; nirodamar, 140. 16; adrodamair, 148. 5; rogknair, 8. 8, etc. ; rornidir, 40. 12 ; of. 178. 20 ; ni ermadair,' 126. 4 ; romenair, 136. 4 (pl. doruimmenatar, 100. 4); tarrasair, 30. 17, 38. 10, 46. 11 (dunarra.stair,* 138. 20); new analogical formatione-conaitigir, 228. 7, ~onaitigair,~230. 17 ; fw~ig&air,~16. 26 ; aroirachair, 68. 21, arr~erachair,~104. 14, 25 ; dellechuir,e 240. 20. Verbs in -ag-, -ig- :- With the exception of coimnigedm ' commemorates,' 136. 18, deponent forms are found only in the third persons of the 8 preterite-rochair$eshzr, 12. 16 ; rocornaicsegestar, 40. 12, ~oc?wmaiccslge8tar,68. 14, roehmfaicceechestar, 252. 3 j rokmai- gestm, 36. 9; rodugalsigestar, 12. 10; r@rgaigestar, 44. 27, cf. 58. 27, 228. 15 ; fothaigesdar, 156. 3, fwothaigestar, 160. 2, 174. 22 (forfothaigeatm-), 194. 4, 214. 13, rofthaigestar, 108. 7, 134. 3 ; coromeglestar, 180. 23 : roimecla~g8itm,44. 26 ; romachtaagsetar, 56. 3.1 Active forms-roardraig, 10.22, cf. 30.2,46.29 ; rocefidaig, 16. 19,25 ; rochmaicsigh, 90. 22 ; rocruthaig, 206.14 ;feidtigit, 90. 15 ; forf&ig, 46. 21 ; fortachtaig (ipv.), 128. 22, fortach- taigj, 220. 21 ; foruashgfe, 42. 13 ; fothaisis, 98. 12, 110. 11, forothaig, 92. 12, 94. 10, 98. 2, 108. 10, 110. 6, eh., rof0thiy, 68, 1, 72. 12, etc.; roimedaig, 128. 2, roimeclaigaet, 92. 6; roingantaigset, 100. 3 ; roinlwchaigset, 126. 11 ; rolagais, 144.7; ronortaig, 16.29; roratkig, 54. 8; rohsdraig, 72. 26 ; roslanakg, 12. 12; ro8oi&g, 42. 8; rosonairtnig, 70. 9; ro8uidt$, 46. 24, 110. 18, 126. 3, roeuidigset, 168. 20; gwtathaig (v.), 252. 18 ; toirthiget, 34. 27.

1 Cf. p. 15, n. I. a Cf. p. 61, n. 2. a From cuintgim ' I ask,' for O.Ir. conaiteeht, Ml. 132d 5, conatecht, LU. 97b I. For eonaitecht is also found, after the analogy of the 8 preterites, m&g, 112. 3 ; cf. conatech, LU. 97a 36. For O.Ir. fmigini (89. 1 huefmgensa, M1. 78d 2). The mtive forwigmi is found 16. 20. 6 Stokes now explains this as standing for *nd-ro-reruchdr, with transition to the dep. inflexion from ad-ro-rerach-, a reduplicated perf. from ad- rec- cognate with con-rkg g. ligo. 6 Cf. Stokes, Indes. Prof. K. Meyer compares deZZgetar, LL. 43b 22. 7 Other deponent forms in the u pret.-rombaituestar, 36. 23, 220. 12; robenukchastar, 106. 27, ef. 162. 23, 150. 16, 210. 5; rocelebraetar, 182, 18; corofermtair, 56. 1 ; roordnestar, 156. 18, 194. 6, 214, 14; ro[i]roiglestar, 68. 32 ; roth&chesfar, 46. 11 : covucsatar, 236. 10. 522 THE DEPONENT VERB IN IRISH-J. STRACHAN.

9. POEXS ASCRIBED TO VARIOUS WRITERS OF ,THE TENTH AND ELEVENTH CENTURIES.

Dallin Mac More (LL. 478 1-49, 478 50-47b 44). Deponents that have become active :- noscechla ('shall hear them'), 47" 12; rachuir, 47" 39; darachuris, 47b 4. Deponent s preterites-rachurestar, 47%20 ; curestar, 47&23; rachloastar, 478 14, 18, 36 (but robris, 478 26 ; rottecht, 47b 10 ; nittuc, 47b 18 ; niratluig, 47b 34, etc.). Verbs in -is- :- cumnigim, 478 41.

Eochaid hua Flaind t984. In the various poems ascribed by O'Curry to this writer I have noted only one deponent form, atamir, LL. 218 26. Cinaed hua ArtacCn 1975 (LU. 51b 13-528 11, LL. 31a 43-32a 34, 150' 26-47, 1548 10-42, 161' 1-1618 43, 161' 44-161' 34, 161' 35-162' 30,' 163' 26-163" 22,' 209' 1-209'' 48 '). Deponent verbs preserved :- rofitir, LL. 1508 35, 161b 19; fetatar, 161b 30; gehair, LU. 51b 32 ; domunemw, LL. 163a 26. Deponent verbs that have become active :- bursit, LL. 1618 37; midit, 1628 5; notmolfat, 2098 30; airisit, 2098 37 ; dodluig,6 20ga 24. Verbs in -ag-, -is-:- corosbrathaig, 16Z8 27 ; rochertaig, 209b 22 ; debthaigin, 209b 6 ; falmaigis, 209&49 ; fotroilsig, 1508 42 ; ciwotlessaig, 2098 14; rorathGig, 209b 3.

1 Poet Laureate of Cerball, king of Leinster, who died, according to the Annals, in the year 908. 2 Ascribed to Cinaed, Book of Ballymote 352. 3 Ascribed to Cinaed by O'Curry, On the dlann. 5. 105; on what grounds he does not say. 4 Cinaed rochertail). cocert, LL. 20gb 22. 6 This seems to he most naturally taken aa a verb, atber frit codian dodluig 6 I will tell you quickly what you ask.' 6 fotroilsiy ' revealed thee, a curious misformation from foillsigiln ' I make manifest,' aa though it were a compound verb beginning with the prep. fo. THE DEPONENT VERB IN IRISH-J. STRACHAN. 523

Deponent 8 pretcrite'-ruosatar, 2098 50 (active forms are common).

Mac Liac tlOl5 (LL. 152s 6-152b 7*, BB. 18g8 36-18gb SO). Deponent verbs preserved :- rofinnathnr ' may he know,' LL. 152b 7. Deponent verbs that have become active :- rachuir, LL. 152a 48. Verbs in -ag-, -6-:- roscwaigset, LL. 1528 22 (rocoiri$idh, BB.) ; 8uidigi8, 152" 23. No deponent a preterites.

Cuan hua Lothchain t1024 (ILL. 33b 10-358 10, 1518 19-1528 4, 199' 35-61, 200' 11-201' 51, BB. 3518 47-352a 23). Deponent verbs that have become active :- roouir, BB. 351b 13; curset, LL. 33b 27; rota~&,~ LL. 348 40 ; rolama, 34 14 ; moltait (raind whom quatrains praise '), 199s 39 ; r~stattlaig,~30 11. Deponent perfecta-romfdair, LL. 199. 50 ; tathamair, 199& 60. Verbs in -up, -ig- :- randdsigset, LL. 15Ib 23; roratha*, 33b 25 ; rowid&~h, BB. 352* 15. Deponent oorwrtoirchestar, BB. 351b 21. Deponent 8 pretentes-&rat8ammlur, LL. 34b 42 ; corchur- aetar, LL. 151" 48 (coralaatar, BB.), roleiceetar, 1518 33 (roth~ecsetar,BIB ), rongiallaadarJ BB. 3528 20. Active forms are much more numerous.

1 In a poem in BB. 81a, ascribed by O'Curry to Ciaed, theee fom are common-roadrastar, romarbaatair, rochuirsedar, rochrechsadar, doluida dar. But I know of no evidence in support of O'Curry's opinion. * Another copy, with considerable variations, BB. 38gb. Also BB. 395. This text differs considerably from LL. and en& at LL. 15Ib 16. 4 Put. of ui@lddur, eid dc thmsa rotacrkl, lit. 'though thy time is young, I will address thee.' 6 If this comes from tothligim, but there is something wanting after the word. 6 Cf. tathaim ' he died,' O'Reilly, LL. 131b 37, 132a 18, 133a 1, dep. rotha- thanauw, LL. 283b 20. 524 THE DEPONENT VERB IN IRISH-J. STRACHAN;

Fland Mainistrech tl056 (LL. 118 19-11b 39, 27b 54-288 49, 131’ 34-132b 4, 132” 6-133” 10, 145” 20-49, 1508 49-150’ 24, 181‘ 1-1868 10). e Deponent verbs preserved :- nijtir, 1828 22. Deponent verbs that have become active :- ni chlune, 1818 42; rochlune, 1848 41; molaim, 185b 42. Deponent perfects-roddmair, 1328 26 ; lamair, 1848 15 ; dmfanair, 182b 53; diashew, 185b 1.’ Verbs in -ag-, -6-:- Wigi8, 1818 29; robethaig, 1328 23; rochetaig, 1818 18; rochechaig, 185a 48; roncruthaig, 133b 4; roderbaig, 145” 34 ; rondorchaig, 185” 10 ; rodubaig, 14Sb 34 ; roiainig, 183b 33 ; ro’olslba;g, 132a 15; rodudig, 1848 10; rodasamaubaig, 145” 32. Deponent s preterites-Of such I have noted only folkastar? 183b 51, from follnaim ‘rule,’ a verb which in 0.h. was deponent. Sctive forms are numerous. Oilla Coemain +I072 (LL. 3b 6-48 46, 1278 1-131b 33). Deponent verbs preserved :- nodtir, 1278 2. Deponent verbs that have become active :- cia cure, 1318 38. Deponent perfects--rogknair, 48 36 ; corogenair, 1318 18 ; genair, 48 18. Verbs in -up, -is- :- romhdaig, 1298 18, 12gb 24, cf. 128b 3; nhdnomnaig, 128b 4. Deponent 8 preterite-~ogabeatar, 12gb 27. Active forms are common.

10. SALTAIBNA RAXUN.’ Deponent verbs preserved :- esaur ‘willeat.’ 1 pl. condessamar, 1266. Seams to be a blunder of the MS., or the facsimile for diardsair ‘ from which grew.’ 2 It is hard to say whether or not the uae of this particular forni was traditional 1n Fland. If not, it is a very curious coinciE:i.in dep* Ed. Stokes; written in the end of the tenth century, as a pears from a reference to a cattle-plague, 1. 2342, in the year 988 (cf. dnnsls of Ulster under the year 986, and the Chronicum Scotorurn under the year 9@), and from the hst of contemporary kings, 4349 8qq. ; cf. Thurneysen, Reo. Celt. Vi. 106, Zimmer, Nennius Yindicatw, 185, 229. THE DEPONENT VERB IN IRISH-J. STBACHAN. 526

fetar. 8.1. ndfetar, 1335, etc.; sg. 3, rojiffr,563, etc.; 8- subj. sg. 1, cofessur, 2883; sg. 2, cofesser, 1327, 6113; pl. 1, ~ofessamar,1265. laimnr. pres. ind. sg. 1. ndamw, 1259. From -ciu the dep. subj. frisaiccidar, 4137, but act. ardasck, 4165. An isolated dep. foichlidar ‘ takes heed ’; of. p. 62, n. 2.

Deponent verbs that have become active :- cluinim. pres. ind. sg. 2. indamchluni, 1182, 1429; subj. pl. 1, cluinem, 3315 ; dep. ipv. sg. 2, cluinte, 1841, 2480, 3339 ; act. &in, 2441, Cnirim. pres. bd. sg. 3. czcires,’ 7241 ; subj. sg. 2 (as ipv.), cuiri, 1561 ; sg. 3, condarcutri, 7827; pret. sg. 3, rotchuir, 1730; fut. sg. l,fochwriub, 6121. foIInaim. pret. sg. 3. rofalhai, 2630. labraim. pres. hd. sg. 3. lakw, 6225. molaim. pret. sg. 3. rommol, 4030 ; romol, 7557. midim. pres. ind. pl. 3. domidet, 99 ; fordamiht, 108. -moinim.a pret. pl. 3. dorumensat, 3689. eamlaim. pres. ind. sg. 1. eamlaim, 375; pret. sg. 3, rossamlai, 1178. -tlllChim. pret. sg. 3. roattlaig, 2593; cf. 3521, 3532, pl. 3, roattlaigset, 3638, 4039. Deponent perfecta-ni chemnacair, 1514 ; ficoemnacar, 1544; nachatamar ‘(since) thou didst not confess,’ 1406;

But read probably guires cdch ‘ that warms all,’ dermoinuir ‘ forget ’ is replaced by dermaitim, na dermaitid, 4869. 526 THE DEPONENT VERB IN IRISH-J. STRACHAN.

rodamair, 7749; duthracair, 5941 ;rogenair, 2245, cf. 2334,3693, etc. ; adassair, 3791, cf. 4791 ; ronaidair, 3121, doroemadair, 2709, 7955 ; cotarrasar, tarrasarsu, 1861, tarrasair, 1633 ; new formations : eon&anaeair, 2798 ; targlammair, 1637. Verbs in -ag-, -1g- :- Deponent forms’ are found only in the 3 pl. of the a pret., roaintadaesetar, 2401 ; romismigsetar, 2409, 3127 ; rosuidegsetar, 4084, by the side of rodentaigset, 2406; ro- rniscnigseet, 5551 ; roauidigset, 5095 (six instances according to Thurneysen, KZ. xxxi. 64). The 3 sg. (about 70 instances) is always active, as are the other persons-roscirugus, 1403 ; f-o~muda~geis,1680, cf. 1320, 1723, 1724; roshraigsem, 1517. Examples of other parts of the verb are pres. ind. sg. 1, nihetrazijm, 8001 ; n.kstecl&Gm, 8002 ; sg. 2, deonaigi, 1431 ; sg. 3, orddaQid, 4873; nachastathigi, 4381 ; pl. 3, bethaigit, 7820: ipv. cwaig, 1597, 2103, cf. 1605, 2115, 2203, 5013, 2877, 6018, 6020, 6021, 6025, 6027, 6031, 6032: fut. sg. 3, Zessaigjd, 2059 ; critfinaigjd, 8056 ; &ra+fes, 4486 ; pl. 3, beccaichjt, 8059.

11. RELIGIOUSTEXTS m TEE Leabhar na h- Uidhre.

DB Br6n flatha Nimc (17&-18).2 genim. raginair 7 geinfes ‘ who has been born and who shall be born,’ 178 1 7-1 8. ScQlaLBi BrBtha (31b-34&). Deponent-ro~Wr,~32a 1,

From other s preterites I have noted the following deponent forms:- rosbiathastar 3413, roemnastar 2767, rooirdnistur 6005, 6541, 7129, rosmach- tastar 1121, ronbaidsemmar 3621, rowaidsemmar 3622, cansatar 4039, roscarsatar 1725, rochomarleicsetar 2737, rodosdairsatar 3665, cf. 8293 ; rodiuhztar 6268, doratsatar 3407, 3508, eondisitar 3771, rofegsatar 4669, &srordiirgsetar 5297, dorigematar 5929, fosrugensatar 5251, cordlaatar 5603, nirlezcsetar 6421, lizcnsatar 5267, roraidsetar 5497, dorinolsutar 2757, tindlsatar 561 7, mesatar 6405, tuesatar 7539. The beginning is wanting in LU. A complete copy, containing a slightly different text, is preserved in LL. 280a-281a. s As to rofoichlitar, 338 ‘14 (uair narofoidlitar ‘became they took no heed’), the ending of the s pet. seems to have been replaced by that of the perf. and the t pret. This would then be an early isolated case of what later became the regular ending. THE DEPONENT VERB IN IRTSH-J. STRACHAN. 527

Active-lamomallat, 32b 17 ; rochmaill, 328 33 ; rochmailset, 33b 29. verbs in -ag-, -is- are always active-pres. connachumnig, 32b 32 ; demnigid, 32b 5 ; demniges, 32b 4 ; lesaiyit, 32b 29 : pret. corocho- macsig, 328 1; rolesaig, 328 4: fut. ordaigfid, 328 11 (bis). Deponent a preterites-doratsamar, 32a 23 ; nadernsarnar, 3Z8 43, in both cases after a preceding atckwmcir.

Sc6Ia na Essergi (348-37b).

Active-atchluinfet, 348 44; tathczdres, 36b 29; genit, 37%31 ; midjid,' 378 40 ; notmolfat, 36b 8. Deponent perfects-dorumiair, 35b 46 ; ragenatdr, 34b 45 ; forodmatdr, 358 17. Transition to the active in forodaim, 35' 22 ; omfodaimsium, 358 24. Verbsin -ag-, -is- are active-pres. ind. demniges, 358 13; cumniges, 358 31 ; chzsnataiges, 35b 5 : subj. sg. 3, coroathnuige, 35b 15; corocumtaige, 35b 11 :ipv. demnig, 37" 29 : pret. racwthraig, 346 4 1 ; rochzcmtaig, 35b 7 : fut. sg. 3, artra&$d, 348 42 ; nosathnzcigfe, 31b 21 ; comalanai@d, 34b 49; cruthaigfea, 39 2; ni erchotiyji, 348 38; fedliyjd, 34b 40, 35b 8, 368 15; arthraigfes, 378 42; toltanaigfees, 35b 3 : pl. 3, notadamraigfet, 36b 8 ; fedZigjt, 34b 8, 3'T8 7 ;fedtigfet, 368 2 I ; dsntadaigjt, 368 43. Deponent a preterite-rothir'riarthstw, 35b 4.

Fis Adamdin (27L3lb).

Active forms-nci cornaillet, 308 11 ; coowend, 2gb 20, tdcurid, 278 4; ni thairiset, 308 31. Deponent perfecdforc&mnacair, 288 42. Verbs in -ag-, -ig- are active-pres. ind. wch&igend, 27b 20; gndthaigee, 3Ib 1; subj. corascoraigea, 27b 32: pret. roartkaig, 27%34; rosc&raig,2gh 1; rofoi~~s~q,2gb 23; roj%o&wg, 31s 14, 25. Deponent a pret.-roinniaemdr, 28b 32.

1 The facsimile has midfed, but the sense requires the fuk--lsand.& midjdsom co$mz fwmadbenib ' then he will pass just judgment on men.' So 34b 9 faidfed stands in the facs., where the sense requires the future. 528 THE DEPONENT VERB IN IRISH-J. STRACHAN.

12. LATERHEROIC TALES. Airec menman Uraird Naic Coisi.’

Deponent forms :- rofetursa, naohjtir, oofessamar ‘that we may know.’ * Deponent verbs that have become active :- arauisbhF3rotamolua, noamolfat. Deponent perfects :- forcaomnacair, commidair, rogenair, roolamair, tarrzcatar. Transition to active in doruimen he thought,’ nimrolam (v.) rodam. Deponent a preterites :- condocorustar, roairmeatar, doruirmestar, rogeisestar, rofait- baistar ; tregMustar, fegustar, cocrastar, scailistar, muainistar, cinneatair-dl in a piece of rhythmical prose ; atraigaetar, rochuitsetar. Verbs in -ag-, -ig- :- dobrethaigsit, rommlaigset, roorkigaet, soilhgiss : ainmniget, brechtnaz$et, bethaigim.

Deaths of Go11 and Garb5 (LL. 107b 22-1119.

Deponent verbs preserved :- nocandgur (v.) 1118 29 ; cofesser (v., subj. sg. 1) 1088 50; atlochw (dom chumachtaib) 11 lb 17.

1 An allegory told to Domnall, king of Ireland, 956-979, by the poet Urard Mac Coisi, when his abode had been plundered by the king’s people; cf. O’Curry, On the Man%. ii. 130 89. I have two copies of the text-one from R.IA. 23 N. 10, which, however, owing to the closing of the Academy, I could not copy to the end ; the other from Rawlinson, B. 512, which I owe to the kindness of Mr. Stokes. The texts are practically the same, only N. supplies some lacunae in R. 2 So, nuder the influence of fetar, awaear, atjmor, atfear ‘ I will tell.’ 3 If this is for aruisfe from arsisaim. R has airsiwa. 4 From fo-&h-tibh mile.' R. has a curious roaitfethair after the -ar perfect. 5 Translated by Stokes, Rev. Celt. xiv. 398 sq., and in part by Zimmer, Zeitsd. f-Deutsch. Alt. xxxu. 208-216. 6 Stokes takes emoradiur ‘ tho’ I speak,’ (v.) 1118 29, as a deponent, but it must rather be regarded as 1 sg. subj. THE DEPONENT VERB IN IRISH-J. STRACKAN. 529

Deponent verbs that have become active I:- ni chlunim 1108 17, 18, 19; menichmalla (v.) llla 32, rfichomaill 1lOa 40; cuir (ipv.) 11Oa 3, 110b 29; nilabair, 108b 7, niroja6air 108b 8. Deponent perfect a:- tarraeair 108b 5. Deponent a preterites :- rothigeetar 108b 49, rottregdastar 108b 47, rohairgtdar 108b 48.3 Verbs in -ag-, 4s- :- nirorathaig 109b 22, conas-athaig 11 la 17.

Bbrama * (LL. 294b-3088). Deponent verbs preserved :- itdgurga (v.) 308b 24, na hdgumar (v.) 308b 17, natdgammar, niagammw5 (v.) 308b 40; rofBtarsa 302b 37,Ptir (v.) 302b 26, Jnta 303b 47 : dmoniursa (v.) 308b 22.s Deponent verbs that have become active :- atohluiniu 3018 18, atchlunim (v.) 305b 29, clunem 308a 34, olui?t (v.) 308b 30 ; wac?wmaill 299b 38 ; mirent (subj.) 300b 28, cuir (ipv.) 3029 27 ; curia 304b 34, rochuir 303b 44,

In 109a 1, nigh dodluig dogasced no dottdrdm eter dkgbco h-Ermn, Stokes sees in dodluig a part of tothhagim ‘l there is none that asks for thy valour, or that counts thee among Erin’s good warriors.” Zimmer translates doubtfully ‘ du hast kein anrecht auf ritterschaft oder ,dass du fiirderhin unter die guten jungen der miinner Irlands gezahlt wirst, but suggests the possibility that dodluig comes from tothlaigim. The parallel passage in the Tdin, LL. 86b 24, connachfuil do dlwig nu dd ddal Iza do dil pi gail laa ru gaisc8d.90 brunsi mbratha, seems to indicate that dluig is a noun with the meaning of ‘claim’ or the like. Cf. LL. 67” 9, fail dlug molta forro, which must mean something like ‘they have a right to praise.’ The noun would naturally be connected with dligim, ‘ I have a claim to.’ Very probably our passage is corrupt, for, apart from the parallel passage in the Tain, we should expect to find after ni not nd but m’. * For 0.h. eia@air we find in this text seiss, lO8b 21, 109‘ 6, 42; cf. Thurneysen, KZ. xxxi. 98 q.; Zimmer, KZ. XXX. 151 q. But it seems hardly safe to infer very much from a form found in eo late a text 88 tb. Why might not sedss have replaced eaassair as narataim has replaced -damair, LL. 1718.4, 8, rogein, rogdnair, LL. 124b 4, etc. 2 The passage in which these forms are found is a variation of a passage in the combat with Ferdiad in the Tdin, LL. 86b 18 84., where the first two forms are also found. Edited by Stokes, Rev. Cdt. Xiii. 32 q. Assigned by Zhmer, Ditsch. f. Deutsch. AZt. xxxii. 307, XXXV. 26, to the eleventh century. The metre shows that in 308b 38 nahagam is a slip for nahdgamar. 8 In 308b 24 is an isolated dep. bri‘gur, probably a formation of the moment for the sake of the jingle with cigar. 530 THE DEPONENT VERB IN IRISH-J. STRACHAN.

darchuhet (v.) 302b 1, corochuirset 295b 10, rochurset 3038 11, chur$mmid 305b 1; md2amat (v.) 306b 18 ; airis ' stay ' 2978 27, 47, 297b 10. Deponent perfect. On n4 ro&ZakJa&, 29gb 4, see below note 5. Actively ni ra$.ulncJiusa' I did not endure,' 307b 3. Deponent 8 preterites :- coromberbastar 300b 32 ; connaigsetw 3048 33, gahsatar (v.) 2998 31, 302" 10, roiadsatar 301a 9, roinnisetw 301b 5, 11, cwdersatar (v.) 2998 34, romhrsetar 295b 6, rordidsetar 295a 39, 302b 19. Yerbs in -ag-, -6- :- nirachtaig 307b 6, rochoraig 3048 42, rochuibsig (v.) 291ib 31, roorddcsig (v.) 300a 50; rddig (ipv. v.) 308% 45, 308b 37.

Togail Troi* (LL. 217L2458). Deponent forms :-atagur 228b 24, atagammar 2208 26 ; rojtir 218&16, etc., rafetamar 240b 23, etc. Active forms :-rachluin 222b 36 ; rmhuir 2188 37, rasirnarchtcir 218a 37, rachuirset 2208 35, 2328 11, curfeetsa 2348 43; samhit 2408 32; airisad 2348 19, cf. 221ib 1, 233b 29. Deponent perfects :--ragenair 2178 36, cf. 217b 20, 2328 21, ragemtar 2178 37, etc.; rolamairk 2278 31; nichczmnacakr 221b 31; dzcthracar 22Sb 18, 20, 3 pl. duthractar 217b 28; cona$mla&gak5 22T8 41, conafmlahgtar 2408 29.

1 In docerddatar 299b 9, rath6cbatar 3018 6, rwgitar 305' 16 we have instances of the Mod. Ir. -adar. 2 A text which has much of the turgid style which became prevalent in later Mid. Ir. literature. In a later portion of LL. (397d-40@) there is a simpler form of the text. I have noted from it atagar 400b 17, rofallnaisit 397b 17, rogeinetar 399s 19, cofeasatar 400* 36, ni rolaln 402a 24, dommnatar 400a 10, tarrasair 406b 9, tavasatur 402b 24, roatlayeatar 39rb 45, 407s 2, caomnacair 402a 53. I have not noted the verbs in -aigim or the deponent 8 preterites. 3 But ni rogePiz 225b 40. Similarly ramideatar 221a 8 for romidair. 4 But PI. niralamatur 227" 36. 5 In 237a 2 nirmlangair has pass. force-nirarlangair do riglnilid na Troianna .i. do Hectair in t-et&alang sain 'that unequal fight was not endurable to the ro a1 warrior of the Trojans, Hector.' Cf. ni rofuladgair do 2eth Cuind 'Conn's Ealf could not smtain (the attack),' LL 299b 4. We seem to have here confusion between the -ar act. pret. and the passive. Cf. LL. 175a 15 niforwlriyither d6uom beith icind intsloig, lit. 'it was not endurable to him to be before the host'; 177a 1 nirbrmbigetlrer [do] Caivpre Niafer eside itir ; also LL. 175a 17 nzi hed barroeblaliyair dosom itir in 'it was not leapt by him at all.' THE DEPONENT VERB IN 1RISH-J. STRACHAN. 531

Verbs in -ag-,-ig- :- Deponent forms are found only in the third persona of the 8 pret.'-raitadaigestar 231a 35, raharthraigaetar 2368 48, 236b 1, rachathaigeetar 241b 42, rachomfaicaigsetar 2258 40, cf. 230b 49, 23gb 40, coracrithnaigsetar 2408 15, rortkaigsetar 2238 19 ; active-raarig 2408 44, rascairig 22Ba 16, 242b 48, rachathaig 224b 34, rach6ra;g 238b 8, rachrechtnaig 224b 40, corachrithnaig 2368 36, oruadaigie 236b 49, rachzcmtaig 223b 16, 17, radmaig 224b 43, cf. 2288 6, 8, oragrddaig 226b 31, cf. 2318 11, raorddaig 234b 47, 48, 2358 I, ras'antalg 2318 38 ; raordaigset 233b 36, cordtreollaigset 2368 39. Other parts of the rerb, pres. ind. deeinn{qinz 2298 2, ar&d 222b 33, mzwlaigid 2408 48, Zuathaigend 243b 5, orddaigit 2238 31.

Cath Ruiss na RigZ (LL. 1718-1788). Deponent verba preserved :- fetar 171a 24, etc.; ratholatharS 1718 3. Deponent verba that have become active-coroscuir 1778 23. Deponent perfects that .have become active-niroataim 17 la 4, 9, 14, nirolarn 1768 27.4 Verbs in -ag-, -is- :- Deponent roredigestar 1768 36 ; rochrithnaigsetar 173b 18, raluamnat&etar 176b 8.5 I have noted nine active forms, all 3 sg. s pret.

Of deponent forms in other s preterites I have noted twenty-two instances of the 3 8 (including rosmolastar), one instance of the 1 pl. rachomraidsemar 231' 6, adtwenty-eight instances of the 3 pl. * Edited by Hogan, Todd Lecture Series, iv. Cf. Zimmer, Zeitaeh.f. Detctach. Alt. xxHii. 219 sq. a Cf. p. 47, ruttolnastair, Rev. Celt. xi. 446, 1. 23. We may have an -ar pet. in rarmetairne 17Ib 2. It may come from a verb "metaim cognate with meta 'a dastard,' BB. 4748 22, 29 (cf. Hogan, p. 6, n.). Of the infixed urn- 'us' for O.Ir. n Hogan quotes other instances. I ma add one more to show how urn- may have spread from the inthitive to the &te verb, LL. 61b 15, Cest didiu cia bad ddig diartiachtain. Ddig iiinartised Chscraid 'Who then would have been likely to have come to 11s ? Is it like1 that Chscraid could have come to us?' Here the inf. ar-tiachtain is foUowJ at once by the finite verb -ur-tised. Other dep. forms in s pret.-raairgestar 1718 41, radercastar 176b 11, 28, cotogdsestar 17Ga 19, cf. 177a 31, rolemstar 177b 18, raliligestar 175. 18, rosceastar li3a 14, rothuairgestar 17'7" 32, tucastar 1776 3; raansatar 176~30, 1W 15, robamatar 176b 8, corog6setar 176a 20, 21, cf. 117a 34, roinnisetar 174: 21, roraidsetar 171a 15, 173a 5, 174%18, ratheigsetar 175* 41. With endmg of perf. and t pret. robenutar 17Gd 35, racuclaigetar 176b 7, rachmm6ratar 177' 10, raergitar 17Zb 4, 1738 24, twxgabtar 173a 51, 176' 37. With regard to this change, e.g. rolommair, rolomsat would easily become rolmwir, rolsminatar on the analogy of tarrmair, tarrasatar, etc., roben, robensat, roben, robenatar after atracht, atrachtatur, ek. Note also cases like rogbetar, where s of the ending is absorbed in 8 of the root. 532 TRE DEPONENT VERB IN IRISH-J. STRACHAN.

Before learing this part of our subject we may juet note briefly the state of the deponent in one or two other late texts, leaving out of sight the deponent E preterite, which has already been illustrated sufficiently. In the Tochmarc Feirbe (LL. 2538--259b, cf. ‘Zimmer, Zeitach. f. Deutsch. Alt. xxxii. 2441, we find niroddmat 255b 1, nid&mair (v.) 2568 9 ; in a fragment of the history of the six ages of the world, LU. 1, 2, rogdnair; in the Book of Invasions, LL. 1-26, only rogknair, genair, tarrasair, by the act. romid (v.), cluinet (v.), corochuir, rochuiraet, acalhat, imacallsat, niiamlaim (v.) ; in the Irish Nennius of Gilla Coemain (LU. 3, 4, BB. 203-211), fetar, tan-raaair, tarrastair, by dorathchuir. In the Leabhar ma g-Ceart (according to Zimmer, Zeitsch. f. BeutscscR. At. xxxv. 11, 89, “friihestens ende des 10 jhs.”), I have noted only the act. midhit. In the Vision of MacConglinne, in the Leabhar Breac version, the composition of which is ascribed by the editor, Dr. Kuno Meyer, to the end of the twelfth century, occur nci fetarasu, duthracur I wish,’ conamtmruaar ‘ I sank,’ by ni condam ‘ did not allow,’ nomisa ‘you will devour me,’ domhnim ‘ I think,’ doairis which consists,’ fuiaidim ‘ I confess,’ atlaigagzs. In 86, 1. 6 midithC lem ‘ I think,’ midithir is used as a passive, cf. atomczcirethar suaa ‘I raise myself up,’ p. 87, 1. 20. In the Story of the Bueid, BB. 449-486, I have noted as dep. nirfidir 4f12~13, fwcmnacair 4568 1, conacamnacw 45ga 36, 47Sa 4, by robuiraed 47@ 27; rochuirset 4818 35 ; rochh.&im 464b 13, rochkiniaiu 16, rochluin 4718 37 j rogenis 45S8 19, genjs 4628 28, genfidh 465b 41 ; roagadl 452b 14, 4568 39, agaillidh 481a 7; labraim 458b 45, labraid 4568 2 j nilamaim 457s 30, lema 4768 42, nirolam 4848 46 j molmaid 466b 30, romolaat 4538 34 ;oirisid 4648 10, etc., rooiris 4638 29, etc. Let us turn briefly to the later cycle of heroic tales, the Ossianio cycle. In the longest text the Acallam na Senzdrach dep. inflexion is found in fetar-rofedarsa, in$dir tu, infedrais, fedamar, fddabair, but in the 8 subj. cofesam. Besides this we find [atlochlamar ‘ we give thanks’ (to God). In this stereotyped usage atlochur, atlochamar survived as a pious expression ; cf. atlochur, LU. 116a 26 (by na Zaim ‘does not dare’), 119b 22, atlochar dzd ‘thanks be to Him,’ Martyrology of Donegal 128, Rea. CeZt. xiv. 48, 1. 14. In the Battle of Ventry I have noted nirlamais ‘ you did not dare,’ 1. 797, but no deponents. In the Macgnimartha Find (Rev. Celt. v. 197-204), cin co fetar 5 15, nifedar 0 25, by nirlam $0 4, 17. In this text occurs a difficult poem, and here we have deponent forms THE DEPONENT VERB IN IRISH-J. STRACHAN. 533

-cuirither saE wan ‘the sea sleeps (cf. suanaid Zer, Gott. GeL An%, 1887, p. 185); tuigither blath inbith (blossom covers the world ’; cuirither iasg mkeoc mbedg ( the speckled fish leaps ’ (cf. folihq id bee bedc, 2.); 6uirither gort glas the green corn rustles ’ (1). This piece was possibly taken from some old text; it has no particular connexion with the story in which it stands.

11. REMARKS ON THE HISTORY OF THE DEPORTENT.

We have now brought together the forms of deponent verbs from a large number of old Irish texts. We have also seen from later texts how verbs originally deponent fared in a later period of the language. The net, it is hoped, has been cast wide enough to take in all the regularly deponent verbs. Doubtless isolated deponent forms still lurk in unexplored corners, particularly specimens of that ill-defined and fluctuating class, the verbs in -ag-, -ig-. But it is altogether improbable that such forms will be found in sdcient numbers to upset materially any results based upon the above collections. It remains to consider what light these collections throw on the history of the deponent verb in Irish. And here it may be remarked that on some points it is as yet impossible to get beyond approximate results. I refer especially to the period of disappearance of the deponent forms. We may arrive at a sort of relative chronology; we see that certain forms disappear earlier than others. We can even fix the time before which the deponent had, except in certain forms, vanished. But, before the limits can be fixed with precision, further investigations in the history of the Irish language will be necessary, for, apart from historical references in the texts, it is on the language that we must chiefly rely for fixing their date. For works of a comparatively late period, such as the poems ascribed to Fland Mainistrech and others, preserved in old manuscripts, the headings in the manuscripts may be accepted without much distrust; but the more ancient the writer is to whom the work is ascribed, the more necessary is it to control these notices by the evidence of the language. The native tendency to antedate may be seen from a glance at Lectures 11. and 111. of O’Curry’s Phil. Trans. 1801-2-5. 35 534 THE DEPONENT VERB IN IRISH-J. STRACHAN. dlannerg and Customs of the Ancient Irish. Meanwhile we will say what seems most probable, distinguishing, as far as possible, between facts and fancies, and trusting to the researches of, let us hope, a not too distant future, to correct errors, and to throw light upon what is yet dark. Even in the present state of the question the deponent will furnish a valuable criterion for testing the antiquity of a text, for it may be laid down as a general principle that the better the deponential inflexion is preserved in a text the older will the text be.’ As we have seen before, the deponent verbs may be divided into two classes. The first class includes radical verbs and a few denominatives, and some verbs which are deponent only in certain parts, usually in the perfect. The second class embraces derivative verbs in -ag-, -is-. These two classes may be most conveniently discussed separately. New developments in the 8 preterite and the subjunctive will be considered afterwards. In the Glosses (see pp. 5-24) the former class includes the following verbs and their compounds dpr ‘Ifear,’ frisailiur ‘expect,’ nirliur,z conairliur ‘ take thought for,’ ‘ consult,’ cluiniur ‘ hear,’ comalmr .‘ fulfil,’ czciriur ‘ put,’ etc., dimeccur ‘ despise,’ fstar ‘ know,’ fociallur ‘ take thought for,’ follziur ‘ fly,’ folnur ‘rule,’ gainiw ‘ am born,’ aclglridzw ‘ address,’ lakur ‘ speak,’ kirniur dare,’ foldrnur ‘ attempt,’ nzidiwr ‘judge,’ -moiniw ‘ think,’ etc., molur ‘ praise,’ condirgur ‘ direct,’ ‘ correct,’ iremidirgur ‘ transfer,’

’ Strangely enough Prof. Zimmer, Ditsch. f. Deutsch. Alt. rsxiii. 269, brings forward as a proof of the high antiquity of the Imram llfaelduin the fact that no deponential forms appear in the pres. system. I have noted in the LU. version only one instance to the point, 1 pl. subj. cuirm, beside the s preterites rolabair, rosamlaisetar, roformdigestar, and the perf. doncnaenair. This are- ment involves the assumption of a pre-deponentid period in historic Irish, at least in the pres. system, an assumption which the preceding pages mill have shown is completely opposed to the facts. As to the dep. forms in the s pret. and the perf., we shall see that the deponent survived there, when it had disappeared from the other parts of the verb. Some of the other proofs of the antiquity of the text are not very convincing. In Fland Nainistrech (t1056) may be found parallels to the perf. forms quoted, such as geguila, cechaiizq, giuil, reraig, rodciuir, corondedaig, to bi, rodmha’ (LL. 1328 10, 20), to the 8 fut. and subj. iarfais (iarfaigim), conkcsem (admad); biru is paralleled by taccu in a poem of Cuan hua Lothchain, LL. 34b 40. So that these arguments hardly prove that the text, in its present form, is particularly old. 2 According to Ascoli, Lex. cxliii. a derivative from airle. 3 A word of uncertain origin ; cf. ,W. dermygu ‘ despise ’ =di-ro-inee-, edmygu ‘honour,’ myged honoratus. 4 Add to the examples molfait ‘ they will praise,’ M1. 69b 1. In other compounds regis regularly active. It may be noted that in these tmo compounds the vcrb has a mental? not a physical, application. Cf. Asc. LPX. THE DEPONENT VERB IK IRISH-J. STRACHAN. 535

refer’ ; samlw ( imitate,’ eechur follow,’ dofuialiur g. labor, -siasiw= hapat (only in compounds), -tluchur (in atluchur (buidi)

( give thanks,’ du-thluchur ‘ demand ’), dtithraccar wish’ ; the isolated admachdur wonder ’ (163, remiaiendider (lo), arahchladar (1I), atmuilniur (19), coinedar,2 immethecrathar, fochehelfatm,s (24) : eaur ded, which supplies the fut. stem of ithim ‘ eat’ ; the pres. subj. of -cIu ‘see’ (23) ; the perfects coneccar, -ddmar, -eiassar (24), cf. tarrastair (22): In the other old texts most of these verbs are found :-dgtgecr (53, 54, 55, 56, etc.), frisailiur (75), airliur, conahliw (65, 59, 69)) cluiniw (46, 48, 52, 56, 58, etc.), comaznw (71, 74), cuiriur (46, 48, 51, 52, etc.), fetar (46, etc., anbj.jfnttar 65, 67, 80), focialZur5 (61), folnur (51, 61, 70), gairaiur (46, 48, 58, 61, 70, etc.), adglddur (46, 48, 52, 54, 55, 61, 64, etc.), Zakur (46, 52, 55, 57, 66), Zaimiur (57, 61, 66, 67, etc.), foldrnw (51, 57, 67, 77), rnidiur (49, 51, 54, 62, 67), -mot%&- (46, 58, 67, etc.), molur (75), samlur (53, 56, 57), -sissiur (49, 54, 56, 57, 58, 62, 70)) -tluchur (47, 49, 58, etc.), dtithraccar (47, 62, 67, 75): eeur.(52, 67, 81); dep. subj. of -ciu (47, 51, 52, 68, etc.). Additional deponents are btiriur roar ’ (50, 53, 59, 69, 70),6 cohrur ‘ help ’ ’ (60, 70), some forms of condim ‘preserve ’ (51, 70),s derdrur ‘roar’ (53), tuigiur ‘cover’ (57, 62, 67, 70, 71 n.). Isolated forms are immusteorathw 68, airlathar (62 n.), bdgur (86 n.), cluradar (59 n.), coiblethar (68), faichlether (621,

1 The dep. forms are found only in the subj. But as only two active forms are found in the Glosses, and these in parts of the verb where the active forms are most eommonl found, this is probably a mere matter of chance. The word is cognate wit1 Lat. satio. 2 In anu nusual conshction, coinedar 2; insalmso ’that David utters this psalm of lamentation,’ where salm is an acc. of the internal object. Cf. expressions like %p+v 8pvov. 3 Cf. note 9. Cf. p. 61 n. 6 Cf. aruacialladar 60, cialtathar 66, ciallastar 69. The verb seems to be a denominative from &all ‘ understanding.’ To dullaim corresponds W. pmyllo considerare, deliberare, Bret. poe2la comprendre, concevoir, to fwhiallaim, W. gobwyllo procurare, curare. Distinct from this verb is fochiallairn ‘ collect,’ 68; cf. Idg. Forsch. ii. 369, Stokes, Urkelt. Sprachtschatz 86. 6 heccesfar (52) is perha s a somewhat slender basis on which to establish a dep. beccwr ‘ roar,’ thougg it may be noted that the word belongs to the eame sphere of meaning aa bpirswr, &hr. 7 I have no exam lea of this verb from the Glosses, except forms where the de . inflexion is not #&inci from the act. Wb. 3* 16, lZb 8. In the Gloasea the verb is active. The verb is usually active; cf. Findisch, Wb. 639 ; Zimmer, Belt. Std i. 72; M1. 90c 15. In Wb. 30d 16 Zimmer takes fochelfatw 88 a dep.= expectabunt. And that seems required by the meaning. Ahin in meaning to fochiallur. 536 THE DEPOXENT VERB IN IRISH-J. STRACHAN. dligethar (68), temadar (47), ronj2adar2 (47), colindfdar (63), luinethar (49 n.), marastar, marathar (56, 'I1 n,), fuinethar (49), mrogatar (65 n.), muinethar (47 n.), condarlaithir (56), conidthar- Lthar (62), rimther (76), rodomsoAtadarZ (47), fossuidiur (71), rantolomar (47, cf. p. 88) ; cf. also conge6ethar (62), ge6ethar (68), fuasnndar (68), immasilzithar (63) and the obscure todonethar, sligethar, nerethar (63). Deponent perfects as in the Glosses docoemnacair (49, 57, cf. 63), forcoemnacair (68, etc.), ddmair (52, 58, etc.), arddmair (58, 68), fordamar (71), siassar (47), awasair (63), tarrasadr (63, 72, etc.). As we shall see, the dep. perf. survived longer than other deponent forms; it even spread by analogy in the later language. I have noted the following forms :-rodsair (dsaim), LU. 528 42, LL. 288a 43 ; of. p. 81 n., rrocesair (Stokes, Lives of Saints, lxxiii.), dosfanair (anaim), immoscoemarcair (oomairckm 49, cf. 63, 71 n.), fkdangair (63, etc.), atacsmnacair, ritacoemfiacair, conarnecar, arlacair, forramair (63), tmacar (65 n.), foruigdnair, conaitigair, arroerachair, dellechuir ( 78),4 tathamair (80), twglammair (tecmaZ- kaim), conchnanacair (83), rarmetairni (83 n), roaitfthair (85 n). Apart from a few sporadic forms, some of which, as we have seen, are probably to be explained as momentary analogical formations, these verbs form a compact and well-defined class. And there are clear indications of the connexion of this class with the Idg. Niddle voice.5 And this connexion will probably

Another example is found in a poem in R.I.A. 23n. 10, p. 20-basi baes fomtemadar etir Zesuib Zalzd. * Borrowed from Lat. do, serutw. In the Glosses scrhtuina is active Wb. 31b 10, Ml. 80c; felardar probably under influence of native temadar. 3 In the Glosses active, and usnaUy elsewhere. 4 In this text fohmadair and folumastar, without any apparent difference of meaning. Stokes takes folumaduir as a present, and such it undoubtedly is in its origin, .cf. p. 13 note, but it may be questioned whether it does not owe its preservation to association wCth the perfects ermadair and the like. 5 Prof. Zimmer, in a well-known paper in KZ. xsx., has tried to show that the deponent in Irish is a purely Irish formation, of a later date than the vowel-syncope produced by the workin of the Irish accent, and consequently later than the introduction of Latin NO& like peccatum, Ir. peccud, gen. pectho. He points out @. 263) that, taking e.g. the dep. cairigedar and the pass. cairigther, cairigther is the regular Irish development of *eariFgitor or the Eke, while cairigedar as the development of the same form is an anomaly. 3ealso (p. 264) urges that the dep. shows irregularly d, while the passive has the regular th. His esplanation is that cairigedar, etc., are late formations built up upon the active mirigid, etc. If that were so, of course any connexion of the IrLsh deponent with the Idg. Middle is out of the question, and we should have to resign ourselves to the strange fact that, some time after the vowel-syncope produced by the Irish accent, for eome mysterious reason these deponent verbs arose in Irish ; that, setting aside the fluctuating class of verbs THE DEPONENT VERB IN IRISH-J. STRACHAN. 535 become clearer when the uses of the middle voice in Indo- Germanic have been investigated. Meanwhile the followhg points may be noted. Many deponent verbs in Irish correspond to middle (or deponent) verbs in other Indo-Germanic languages (cf. Thurneysen, Idg. Forsch. i. 462). Thus sechur corresponds to Lat. sepor, Gr. ZTO~QA, Skr. ~(icate, Zd. liacaite; gainiur corresponds to Skr, jiyate, Zd. %ayeit?, Gr. +ppac (a formation corresponding, except in the grade of the root, to the Irish fut. adgignethw, p. Sl), Lat. nascor; -moiniw corresponds to Skr. m&?zyate, Zd. rnatay& cf. Lat. -mitaiscor. So -tluchur has been compared by Stokes with Lat. Zopr for Wopuor, -glhdur with Skr. h.rcidaat.9. Further, dgur (perhaps with the vocalism of the in -gur, the deponent inflexion established itself in a small and well-defined class of verbs, some of which correspond to deponent forms in Lat., and others to middle verbs in Greek and Sanskrit, and that this new formation had begun to ass again into the active in the earliest Glosses, and before the eleventh century ha$ except in certain forms, practically vanished. But before we give credence to this strange tale, it would be well to examine the facts, and see whether no probable ex lanation offers itself. Zimmer's statement that a form like eairigedar, if 08,is not in accordance with the laws of the Irish accent must simply be accepted. His other statement that the deponent shows d where the passive shows th is only partly true. The ending -dar is the usual one in deponent verbs in -yur, though even here we find -thar in adamrigethar, erbirigithir. But in the other class of deponents, as a glance at the lists will show, -thar is much more common than -dar. Now, on Zimmer's theory, . this -thr is as difticult of explanation aa -dar is easy. Thurneysen, KZ. xxxi. 63, has suggested a way of avoiding the inference that Zimmer draws from forms like suidigidir. The distinction between dep. szdidigickir and pass. suidigtMr is a secondary one, due to the desire to distinguish deponent from passive forms. Active forms like ainmnigid, hnigid might easily produce in place of the re,dar *suidigthir a form suidigidir, which then drew after it a conjunct suidigedar instead of *suidigthw. For this claas Thurneysen's explanation is very probable. But how about the other class of deponents, where there were no parallel active forms? The difference of ending -thar, -dar forbids us to see in them analogical forms after verbs in -gw. Besides, a8 we see from a comparison of the Glosses with the profane texts, these verbs in -gur must have been to a very great extent learned formations, though, of course, the type must have been there before. Why then do we hdZaimethar instead of *Zaimthw from lamitor, -cialZathar instead of *ciaEllar from *ceiZkitor? I would suggest the following explanation. From cdmlanktm would come *com$atar, comahatar. This should have given 'ccwnalntar. If this form is to be assumed as an intermediate stage, n must have had the value of a vowel, whence *eomaZnatar, comalmthar (cf. the pass. foircnither from fdrea&tor). The passive is here distinguished by a change of conjugation, comakither. So in the pl. cdmlancintor would give comalnatar. Similarly "lubvcitdr would gi-re first 'labgtar. Under the influence of Zabrur g might here develop not to ar but to va, whence Zubrathar, and in the pl. hbratar. A similar explanation would apply to airlethar, conairlethar, tonwathar, tomnatar, insamlathar, disamlathar, dofuisledar. Since the Irish were evidently resolved at any coat to keep the deponent idexion apart from the passive, a ready enough means of doing so lay to hand by simply extending the endings -athar, -&hat-, -atar, -etar throughout the deponent-gaiaethar : gainiuv=airZethar : airliw, etc. In 538 TH~DEPONENT VERB IN IRISH-J. STRACHAN.

perfect, Brugmann, Grundriss, ii. 1256) may be compared with Gr. ~XO~UL,midiur, perf. tnbdar, with 8r. ,&opur, ,u+opac, though the formation of the present stem is different ; with the forms from this verb meaning to measure’ cf. Lat. metior, Skr. mimite. Ir. -8isaiar with its intransitive signification suggests comparison with the Gr. middle TUTU^. Vedic Sanskrit has both tk%t@i and tishthate in a neuter sense. But this use of the pres. act. is probably less original than that of the Gr. LT~~L,compared with ZUT~K~,&qv. For the usage in Vedic Sanskrit see Delbriick, Attindische Sintax, 257 sq. Xiathar suggests at once Lat. moror. The Glosses certainly have the active, and that is so far against the direct comparison of the two deponents. On the other hand, the language of the Laws is highly archaic, and it is not impossible that an old deponent survived there, which had become active in the Glosses. The root He?, which elsewhere is usually active (but cf. Lat. clueor by clueo), is in Irish deponent, except in the perfect.’ Root ed- ‘eat’ appears only in the 8 formation ed-s- and as a deponent; cf. Gr. 280p~~ and isolated forms of Skr. ad- (Delbriick, AZtind. Synt. 233). In some cases in the passive itself a similar secondary distinction has been introduced. Thus to-ilkit& and to-dZlint?r would have become do-elltar. Perspicuity demanded a separate plnr. form, and we have doellatar, though the ending ’ -atar, probably under the influence of the active -at, spread further than was absolute1 necessq. A5 to the d in moladar, etc., whether it is to be explained?, due to Agedar or in some other way, it is also found in the passive, e.g. micsedar. It was onl in the third person that the regular developments of the deponent could colliie with the passive. In the other persons it was free to go its own way. Thus, in 1 pl. *coumiliggimm became coumiligmmer, *dechrigimor dechrtgmir, in 2 sg. *miditer became mitter, *maniter naointer, *carigiter cairigther. It may be noted that in Class I. a vowel is regularly found before the heavy m in the 1 ll., ueohemmar by cosmiligmm. This difference seems to depend upon the di erent number of syllables. It is only necessar to refer briefly to a couple of Prof. Zimmer’s other assertions. On 269 Ee says, “im praseas dagegen ist deponenUexion bei wnrzelverben $:h. bei pr&ensstiimmen auf o (Lat. lego-, dico-) ausserst selten und aktive formen liegen daneben.” How far this is true wiU be seen from a glance at the forms from the radical verbs clviniuv, gainiur, midiwr, -moiniwr, sechur, -sissiw. Ba to the assertion that we can still, in O.Ir., see the deponent pwing, the above lists confirm Thurne en’s remarks, KZ. xxxi. 63. A urther indication of the connexion of tc Irish deponent with the middle voice is the ending -the, -ther of the 2 sg. (Thurne en, Idg. Forsch. i. 460 sq.). In the 2 pl. Ir. uses the active ending where gt. has the new formation -mini. This was probably to avoid coincidence with 3 sg. In the pres. sec. r endings are wanting in the dep. as in the pass. Here an -adar (-athar), -e&r wodd in most case8 have been identical with the conjunct form of the present. 1 Is the prevalence of the de onent inflexion in Irish due to the influence of laamr ‘Bpeak’? On the otler hand it looks as though IdF. *&mwmi had become in Irish *cZvni&(r) under the influence of *cesi5 ‘888. The active perf. chakz compared with the dep. erfects gdnair, &air, midair indicates that the dep. inflexion in Irish is secon&ry. THE DEPONENT VERB IN IRISH-J. SPRACHAN. 539 dke8 ‘ see- ’ (cf. Skr. ccieh.te cdkshate) only the present subjunctive is deponent. Condim (protect’ (Skr. dvati) is, as in Skr., active in the Glosses, but shows deponent forms in some old texts. Instances of verbs middle in some parts and active in others are found in other languages, of. Delbriick, Altind. Synt. 235, Gr. pahw P~OO~UL,etc.; in Greek it is particularly in the fut. that middle forms are found. Let US glance briefly at the meaning of these deponents. One well-defined group denotes the exercise of the organs of speech-

-tluchur,’ labrur ‘ speak,’ ad-glddur ( address,’ arafocladw, p. 11, atmuilniur iterum dico, buriur, derdrur ( roar.’ With these may go molur ‘praise,’ if, as seems likely, the sense of praising has been developed from that of speaking.2 Cf. Lat. verbs like wocifeor, pzdhitor, Skr. verbs like bhhshate ‘ speaks,’ Greek verbs like ~~VXLO~UL,p~iop~~. Other verbs denoting the exercise of a bodily function are -tear (see,’ esur, further follhur gfly’ (cf. Skr. prdvate). Of the other deponents it may be noted how many belong to the mental sphere (cf. the Sanskrit verbs, Delbriick, Altind. Synt. 233)-dgur ‘ fear,’ frisailiur ‘ expect,’ comalnur ( fulfil ’ (but Enaim ‘ fill ’), condirgur ‘ direct,’ dinaeccwr ( despise,’fetar ‘ know,’ lnimur ( dare,’ foldmur ‘ attempt ’ (cf. Lat. copaor), rnidiw ‘judge,’ -moiniur ( think,’ -8amlw imitor, dhthraecar ‘ wish,’ tolur ‘ please,’ folnur ‘ rule,’ airliur, fochiallwr (take thought for.’ Akin in meaning to the last verbs is cobrur ‘help’ (cf. Lat. auxilior), with which go the deponent forms of condina and temadm ‘may he protect.’ With these may be mentioned fosuidiur, which in the Laws is translated ( entertain ’ ; cf. fosughadh ( maintenance,’ O’Donovan Suppl. In cuiriurs we often find an intransitive meaning (p. 8 n.), or a middle force ‘take to oneself,’ but in some passages, e.g. M1. 52, the meaning seems to be simply actire. It may be noted that the active forms in the Glosses (p. 9) have an active force, so diohuirid, imdacuiret (p. 56). As to dojkisliur (Lat. salio), dufuislim is found glossing Zabo,

1 In atluchur becidi ‘I give thanks.’ In do-tluchur it has rather the 8enm of asking for oneself, akGrt7ar. 2 Cf. p. 19 n., alao Wb. 14b 26, qua et tomoltid armbdis ‘praying and calling for (P) our death,’ 20d 4 aimthomoldid ‘ do not call upon me ’ (?), where see Stokes’ note, cf. KZ. xxxi. 240. s With this I would put immthecrathar ‘ covers’ =*imme-do-aith-corethar. This verb may have influenced -twigiew ‘cover,’ which was also exposed to the influence of derivatives in -giur. 540 THE DEPONENT VBRB IN IRISH-J. STRACHAN. dofuisledar, etc., glosses labor, elabor. The other verbs call for no further remark. With regard to the perfects cotanecar, forodamar, adrodamar, aiassair (from dsed->, it is hard to say whether they come from Idg. middle forms, or whether they are the beginning of the analogical extension of the -ar perfect which is found in later Irish. If the variation between active and deponent here be old, we may compare Lat. gaudeo, gavisus sum and the like. Of deponent forms related to active verbs as the Greek and Sanskrit middle to the active the examples are few and not very certain. I have already noticed coinedar, Nl. ’14b 1, with its peculiar construction. In LU. llOb 31 (p. 68), Windisch takes nosinethar (probably from no-n-sinethar with infixed pronoun) as deponent ‘he stretches himself.’ The regular passive to shim ‘I stretch’ would be sinter. So imrnashithar ddib (p. 63 n. 2, of. p. 68 n. 7) lit. ‘there are mutual stretchings to them.’ For the construction cf. Wind. Wb. 415, o’manacca d&b ‘when they saw one another,’ LL. 256a 37, and, from a deponent, imasisedar doib, p. 54 n. 4. In LU. 1068 44 (p. 68), Windisch takes fhasnadar as a deponent form from ftiasnaina, and the reflexive meaning suits the passage well j nosfuasnaither im Conculaind, Lu. 12P 36, seems certainly passive ; cf. the passive construction of *~?&saim, Wind. S.V. da8tw, Ml. 50d 2, LU. 638 8. In the Glosses and usually in the profane literature bertaigim ‘ shake ’ is active. On p. 53 we have nombertaigedar, etc., ‘he shakes himself.’ I may refer also to colzge6ethar p. 62 n. 3, gdethar p. 68 n. 5. Perhaps an examination of other old texts will bring more examples to light. The second class of deponents, the denominatives in -giur, resembles, in respect of the freedom of its formation, the various Latin denominatives with the suffix +o- (Brugmann, Grundr. ii. §§777-8). It is a formation common to the two branches of Keltic, cf. Ir. aciraigur with W. snrhaf, though of course deponent forms are found only in Irish. In Irish it spread most widely in the learned language, the language of the church, and it is found in the largest numbers in the Old Irish Glosses. In the great majority of cases there the verbs are evidently new formations coined after the stamp of the Latin words. In some instances the glossator makes this clear enough. In Nl. 128~8 tutabitur is glossed by inilligfid .i. doeinfea : here initligjd is coined THE DEPONENT VERB IN IRISH-J. STRACHAN. 541 to express ktabitw ; doernfa is the usual Irish for will protect.’ Similarly in MI. 30c 18 potahnt is glossed by deecgaejfit .i. ibait 8on. As might be expected, the formation is not very frequent in the old heroic tales. In some of the simpler of them, such as the Tdin Bd Regemain and the TGin B6 Regamna, it is not found at all. In the comparatively long Tdin B6 FrGich are found only forms from airigiur, which in the Glosses is regularly deponent. A comparison of the older LU. version of the %in Bd CuaiZGge with the later more literaq version in LL. in respect of the frequency of verbs in -is- is ,instructive. In later ecclesiastical texts like the Tripartite Life of 8. Patrick, the Saltair na Rann, and the Passions and Homilies edited by Atkinson from the Leabhar Breac these verbs swarm. From the language of the church they must gradually have made their way into the language of the people; in Mod. Ir. the formation is very numerous. It would be a vain task to try to enumerate the verbs of this class, for each newly discovered text will probably add new forms. Thus the Begula of Columbcille furnishes the forms nomuinichit-,l coscelaigetar, the Arrada mani cumrigedar, the metrical treatises edited by Thurneysen, Irische Texte iii. 1 sp., oentadaigetar by deochraigetar.z But it may be of interest to arrange the forms found in the three great collections of Glosses, to see how far the deponent and how far the active inflexion prevails in different parts of the verb.

Pres. ind. sg. 1. Deponent- Wb. naadamrugw, niconairigur8a, nochairigur, fiomistigur ; Ml. coasmecnugursa,3 nothoris- nigiur.

1 Read nomuimichither-cacB bec n6 cech m6r domuinichether. Reeves translates ‘thou possessest.’ I have no further examples of the word. 2 It is interesting to note how the deponent forms are found in the questions, while the answers show active forms, e.g. 3, el na soerbaii-d d&, cid nosdeocrai- gethar ingradaib? Niansa : as alidanaib forhat 7 deocruigid. It looks as though the questions had become formulae in which the old forms survived. Exceptions to the rule are, however, found in the version on p. 24. In the rammatical tract, BB. 314 q.,the text has the dep. condadeliggitar, 323. 26 Ey other deponents like aosdeachmthar distinguishes them (=da’-corethar) 31gb 5 ; the comm. has the active fograiyit, fail2sige8, etc., but also nasdeligidar 320a 37. 3 Cf. p. 40 note. At the game time the syntax is decidedly in favour of the subjunctive, for the meaning is ‘that I should root out,’ not ‘ that I do root out,’ so that I should be inclined to leave it an open question whether we have not here an early instance of the subj. in -to‘. The other instance, mothvrisnigiur, seems to be indicative, but we have as yet no investigation into the uses of the moods in Irish. In the absence of such an investigation, I fear I may have in some cases classed indicatives as subjunctives, and vice oelrd. 542 THE DEPONENT VERB IN IRISH-3. STRACHAN.

Active-Wb. conasarcagim, sulbi[rigim] ; Sg. bruthnazjrim, cumachta+im, mmeigim, dainglaigim, nidephthigim, dechrigim, imtrknigim, citgenaigim, athrigim, baithaigim, cuiligim, failligim, fescrigim, jiuchaigim, gaigim, gortigim, intonnaigim, mzinzjrirn, rkchtaigim, trebairigim, ualligim ; MI. tosigachtaigim. sg. 2. Deponent only-Tb. nocairigther ; Sg. cumachtaigther j MI. adbartaigther, dimigther. sg. 3. Deponent-Wb. adamrigethar, beo@idir, nombeoigedar, cairigedar, cuimniyedar, dixnigedw, fedligedar, jirianigedar (bis), fogrigedar, foirbthigedar, toirthigedar ; Sg. deligedar (bis), deaimrechtaigedar, disruthaigedar, dixnigedar, comhi- cnz$edar, engraicigidir (bis), engracigedar, etarcnigedar, ilaigedar (bis), oenaigedar, sainigedar, -suidigedar (bis), trebrigedar ; Nl. adamrigedar, adbartaigedar, adeitchethar (ter), aidlicnigedar, adribaigedar, iruthnaigedwr, cluehi- gedar (bis), cobrigedar, comadasaigidir (bis), cruthaigedar, danaigediar], dimigedar, kcmailtigidir, erbirigithir, erdar- caigidk, erladaigidir, erladaigedar, etuailngigedar, nosess- raesaigedar, fercaigekr, jiachaigedar, foilsigidir (bis), foilsigedar (ter), follaigedar, -fortachtaigedw, guaigedar, imdaigidir (ter), lobraigedar (bis), londaigedar (bis), mothaigedar, rediyedar, eon&aaigedar, suba@idir, -tesaigedar, immetrenaigedar, -tuailngigedar. Active-Wb. dlgenigid ; Sg. demnigid, ailigid; Ill. erbirigid, imdaigcid], nosizaigid. pl. 1. Deponent - Wb. adiilgnigmar ; Sg. cosmiligmmer ; MI. dechrGmir. Active-Wb. foruaisligem ; MI. denznignii, elithrigmi, failtigmi. pl. 3. Deponent - Wb. holtigetar, dixnigetar, -6ttigetar ; Sg. aidlignigitir, aidlicnigetar, bindigeddar, cosmailigetar, dechrigetar (bis), engraicigetar (bis), -fai&getar (bis) ; MI. adamra~etar,cluichigetar, dechraigetar (bis), condo- gailsegetar, eclidigitir, -Jndbadaigetar, guaigitir, inadaigitir (bis), imdaigetar, lobraigetar, mindchigitir, suidigitir, huatigidir. Active-N1. subaigit, -feidliget, cojnbualzaigit, -ai&get; relative Wb. iudigte (19a 18), Sg. foilaigdde; Nl. cathaigte,

1 Both Stokes and Zimmer read swlbi mgim], but it seems B ually osaible, and it would be more in accordance with tL e usage in Wb., to real sulbi&gur]. THE DEPONENT VERB IN IRISH-J. STRACHAN. 543

dechaigte, demnigte, feidligte, fograigte, greschaigte, mencigite, saith-aigte, semigte (86a lo), taraisnigte, acarbaigte, aithirgigte, intbdaigte, lainnigte, menaigte, sitaigte. subj. eg. 1. Deponent only-Wb. rofoirbthiger j M.1. coerladaig:ar, nufailtiger, nosilaiger. sg. 2. Deponent-Wb. cairigther j M1. adnamraigther, danaigther, ktaigtgtilw (bis), fotchridigther, inraiccaigther. Active-Ml. -daingnige, -imdaigi. sg. 3. Deponent-Wb. dfxhrigedw; Sg. senaigidir ; MI. iresmini- gedar, friscathaigedar, comaicaigedar, comdemnigedar, cutrummaigidir, diummussaigedar, -dixnigedar, -etwcni- gedar, focridigedar, grkschigidir, lobraigedw, fmmamaige- dar, moithaigidir, semigidir, sochenelaigidir, sonartnaigedar, uraigedw. Active-MI. GO adbwtaigid, GO mothaigid, co thirmaigid, mani mesraigea. pl. 1. Deponent-Wb. fedligmer. Active-Ml. nuallaigem (cf. boltigme, Acr.). pl. 3. Deponent - Wb. inbothigetar; MI. adamraigetar (bis), failtiyetar, inlinaigetar. Active-Wb. herassiget. ipv. sg. 2. Deponent only-MI. aditchide, feronigthe, fochridigtk, foilsigthe, trebrigthe, dianaigthe, ollaigthe, ruccaigthe (270 ll), sldnaigthe. s pret. sg. 1. Deponent only-Wb. -airigsiur (bis), roardrigestw, rocathichsiur ; M1. roadbartaigsiur (bis), rodoichemlaigsiur, rufrtythanaigisiur, rusuidigsiur. sg. 2. Deponent only-Mi. ruceataigser, rofoerbthichser, rotethnaigser, rosudigser, rosuthchaigser, rotaitnzgser. sg. 3. Deponent only-Wb. roardrigestar, roddnaigestar, radeem- nigestar, rafrianigestar, rosfailszgestar, romiscsigestar, roordigestar (bis), rosuidigestar ; Sg. cichnaigistir, rotol- tanaigestar j Nl. roadbartaigastar, roainmnigestar (ter), rosoosmailige~tar, rucuhgeatar, racumgaigestar, rooutrum- tnaigestar, rodainglzigestar, roddnaigestar (bis), rodumaigestar, roetaronaigestar, rue'tuailng[ig]estar, rofasa~estar, rondfrinnaigestar, rofoilsigestar (quater), rolondaigestar, fosroamrnamigestar, rondoirammaigestar, romincigestar, radmrechtnigestar, rusaithaigestar, rwarigestw, rosonart- naigestar, rosuiigeetar (ter), rutessaigestar, rotorasnaigestar, rothuailngt&dar, runuaibrigestar. 6.14 THE DEPONENT VERB 1N IRISH-J. STRACKAN. pl. 1. Deponent-Sg. rochruthaigsemmar ; Ml, rufeeidligsemmnr. Active-M1. rondoirmaigssm, roerlGr@em. pl. 3. D eponent-W b . roirladigsetar, rofoilsgsetar, forrusuidigse t ar ; Ml. rodebthaichsetar, roeehtrannaigsetar, rognathaigsetar. Active-%%. rdsdrichset ; Ml. adruamraigset, road6 ffirtaigset, rofidligset, rufdilsbset, robertaigset, ronddmainnigset, roe'trummaigset, rofeszdchrffiigset,rumadaigset (bis). fut. sg. 1. Deponent-Vb. gaimigfer ; MI. adbartaigfer ; (cf. Acr. dthtjfar). Active '-MI.. -etarcugube, greachaigfeffi. sg. 2. Deponent-M1. mescaigjder. Active-Ml. ad6artaigfe, ilaigfi, -daingnichfe. sg. 3. Active-MI. comaicsigjd, ddnaigfea, addanaigfea, feidligjd, foilsigjd, imdaigjd, suidigjth, arsada@th, e'trummaigjd, inikligjd, erdarcaigfes. pl. 3. Active-MI. cujnachtaigfit, dechraigfet, treni;feidhgfet, deugaigjt.

If we compare these different sets of Glosses with one another, we see clearly a fallingoff in the deponent inflexion and an increase in the active. In the oldest of the three collections of Glosses, the Wiirzburg Glosses, active forms are few; in the Milan Glosses, even if we make allowance for the greater number of instances, the proportion of active forms is greater. But it would hardly be safe to reason backwards, and assume that at one time these verbs followed the deponent inflexion entirely, for, as we have seen, the influwce of the active forms is probably to be seen in the development of deponent forms like suidigidir, suidigitir. The proportion of active forms to deponent differs in different parts of the verb. Some parts, e.g. the singular of the s preterite and the 2 sg. imperative, show only deponent forms. In the future, on the other hand, except in the 1 sg., and one instance in the 2 sg., the forms are all active. But it must be borne in mind that the examples come from the Milan Glosses. Another thing that may be noted is the prevalence in Sg. of the active inflexion in the 1 sg. pres. ind. It is true that most of the examples are imitations of Latin words, and that many of them are found only in this one form, but eyen then there is no reason

* In TVb. 12d 12 Stokes and Zimmer extend 8211. to sulbaivigfeea. THE DEPONENT VERB IN IRlSH-J. STRACHAN. 545 why the writer should have used the active form, had it not been more familiar to him in this part of the verb. Unfortunately MI. has very few examples of this person. But the same thing is seen in the former class of verbs--frithlim MI., demeccim Sg., dqiuislim Sg., duthluchina Ml. Wb. has here always the deponent form. And in the old secular texts that I have examined the deponent inflexion prevails-atamidiur (p. 51), adgladur (55), rocluiniur (56), domuiniur (58), cluniur (59), etc., but rotchlunim in a religious text (p. 72). In the other old texts it is difficult to compare the inflexion of the -is- verbs with their inflexion in the Glosses, both because of the rarity of their occurrence, and because such instances as are found come almost entirely from the third persons of the present and the s preterite. A 1 sg. pres. ind. noohosnagur is found p. 47, a 2 sg. -muinichkther p. 98. Of the 2 sg. subj. we have examples in fuiligthw p. 75, and mebraigthr p. 76, by the side of the active mani bretha$eseo p. 68. Of the 1 pl. pres. ind. active forms are found in cathaigmit p. 68, cdraigmit, sidaigmit p. 64 ; dep. airigmer p. 52. Of the 1 pl. pres. subj. an active form occuv in aithigm p. 75. In the 2 sg. ipv. I have noted only the active inflexion, ardotehuibdig p. 66, in a text where the active inflexion prevails in these verbs ; similarly in the future 1 sg. brethaigjt p. 68, 3 pl. crechtnaigjt p. 71. In the 3 sg. of the a preterite deponent forms are common, less frequent in the 3 pl. But active forms are not rare-aamaiges p. 47, rof;cachtn&g p. 53, ratha@ees p. 57, rachasaig p. 64, etc. This represents a later state of matters than in the Glosses. The evidence indicates that the deponent inflexion passed into the active at an earlier period in this class than in the other. The Milan Glosses clearly point in this direction; so do the old profane texts, unless indeed we are to assume that time and the scribes have dealt more hardly with these verbs than with the others. In the Regula Mochuta, too, p. 73, where deponent forms would have been protected by the metre, are found the active nonmiscnegat and aithigem. A11 these facts are in favour of the above conclusion. As this class of verbs is found chiefly in the Glosses, and as it is not largely represented in our other old texts, it has seemed most convenient to treat of it briefly as a whole, before we proceed to remark on the history of the deponent generally in Irish. As in the deponents in -giur so in the others the deponent idexion has already in the Old Irish Glosses begun to give place to the active. This change has naturally gone furthest in the 546 THE DEPONENT VERB IN IRISH-J. STRACHAN. latest of the three great collections of Glosses, the Milan Glossee? Active forms are found most frequently in comalnur LfulfiL' Of active forms in the 1 sg. pres. ind. I have already spoken. The other active forms are distributed as follows-pres. ind. sg. 3 comallaid Ml., comalnas Wb., labraid Ml,; pl. 1 ZaBraimnas m., duntlucham Wb. ; pl. 3 comahat Wb., comallait MI., ckmallaite Ml., immusaccaldat Ml., molat Ml. :-subj. sg. 1 fristacw MI. ; Sg. 2 cuiri Bcr., intamlas Ml. ; sg. 3 comalla M1. ; pl. 1 kkaimme M1. ; pl. 3 chomalnit Wb., dichet Ml., molait Ml. :--8 pret. sg 1 rocomallus M1. ; pl. 3 rochomalnisset Wb., wufolum~atMI., foruraithminset Ml., arrumuinset Ml. :-reduplicated future only deponent cechlallnr Wb., MI.:--- fut. and subj. Bg. 1 doduthrk Wb.; sg. 2 dhthais Cam. Ml. ; pl. 3 imroimset, imrzcimset Ml., dunfutharset M1.4 f~t.~sg. 2 folnaibe Ml., dzcndamro~mneJ%seM1. ; sg. 3 dekntamlafa MI., nicontuialijea M1. ; pl. 1 atluchfam Wb. ; pl. 3 comallaibte MI., mdfait M1. The 2 sg. pres. ind. and pret., the 2 sg. ipv., and the 3 sg. pret. are always deponent. If we turn to the other deponent texts, we find in some of them the deponential Mexion preserved as well as, if not better than, in the Glosses, while in others the active inflexion has made greater inroads. This encroachment of the active inflexion might 'be explained in two ways. Either it represents the fluctuation of the language at the time of the redaction of the

According to Thurneysen, Rev. Celt. vi. 318, the manuscript in which the Milan Glosses are found belongs to the end of the eighth or the beginning of the ninth century. From a comparison of the language with the Irish of the Book of Armagh, on which however see now Zimmer, Zeitsch. f. deutsch. AZterth. xxxv. 53, 78, he would place the original at the latest in the middle of the eighth century. It is clear from the scribal errors that the glosses have been copied once, if not oftener. The Wiirzburg Glosses are preserved in a manuscript of the end of the ninth or the beginning of the tenth century, but, as Thurneysen points out, the language is considerabl older than that of the Milan Glosses. He would put the original in the seventi . The St. Gall Glosses, preserved in a manuscript written in 845 or 2. xxxiii. 93 note, he suggests, lie between the two. Of course this datin is only approximate. It were greatly to be desired that some one well verse1 in the theology of- the time should examine the subject-matter of the Wiirzburg and the Milan Glosses. Therein, so far aa I see, lies the only hope of fixing more accurately the date of composition. It would be interesting to know, for example, if there be any trace of the influence of Isidore on the Wiirzbnr Glosses. A superior limit for the data of the Glosses on Bede is fixed by tie date of the original ; unfortunately these glosaes are scanty. * With transition to the 8 pret. s The 1 sg. fut. is always deponent ; of the 2 sg. no deponent forms occur in this class of verbs, but it is to be noted that the only two examples are from Ml. To the dep. forms in sg. 3 should be added cwna'laabadar, MI. 46c 20. THE DEPONENT VERB IN IRISH-J. STRACHAN. 547 text, or active forms have been at a later time substituted by copyists for deponential. Of the latter examples are not wanting. Thus, on p. 51 cotnofadar is evidently more original than cotndba.' On p. 54, for adglaasnaarni, of the Yellow Book of Lecan, Egerton 1782 has atroglksimni, and H. 1. 13, T.C.D., which agrees closely with Egerton, has atrogleaimni. On p. 69, for acailli of LU. and H. 1. 13 the Yellow Book of Lecan has the deponent adgzaiter. But such cases are few compared with the huge mass of instances in which the deponential forms have been faithfully preserved. I may say that I started with a prejudice against these Irish scribes -so many are the sins that have been laid to their charge,-and was prepared to find the old deponential inflexion freely altered in later manuscripts. But, as the investigation went on, I was more and more shuck by the fidelity With which these old forms at least were copied. Thus, the later Yellow Book of Lecan furnishes texts in which the deponent inflexion is every whit as well preserved as in the earlier Leabhar na h-Uidhre or the Book of Leinster. In ordel: to subject this point to as severe a test as possible, during a visit to Dublin I collated Tain Bo Prdich (p. 52), Longes Nac n-Usn@ (p. 53), Tdin Bd Regemaha and Tiin Bd Regamna (p. 54), and Echtra Condla (p. 69), with copies in the modern paper manuscript, H. 1. 13, in Trinity College. The result was a triumph for the fidelity of the scribe. I take as an example the text Echtra CondZa (p. 69) ; the corresponding forms in H. 1. 13 are, in the same order, totcuirether, docorustuir, accalli, adgladadm, coto- megnigstar (=cotomeicfiigetar Yellow Book of Lecan). The conclusion has forced itself upon me that where active forms are frequent, or where the deponent inflexion has practically given place to the active, we hare to deal with something more than a mere copyist's alterations. But to this we shall have to return again. In the texts of which we have been speaking certain parts of the verb are represented by numerous examples, particularly the third persons of the present indicative and subjunctive, the perfect, and the 8 preterite. It would be useless to repeat all

1 If we should not here speak rather of the redactor than of the scribe. There are clearly seyeral redactions of this old text (cf. p. 50 n.), of which the LL. version is by no meam the best. 2 The Tiin Bd Regemain stands on pp. 346-7, the Tain B6 Regamlza on p. 346. The text is the same as in Egerton 1782. 548 THE DEPONENT VERB IN IRISH-J. STRACHAN. these forms again. It is sufficient to note the forms of less frequent occurrence. We have already remarked that here the 1 sg. pres. ind. is usually deponent :- attdgur (52, etc.), frisailiur (75), cluniur (56, 59), tocuiriur (46, cf. 60), adglddur (71, 55), Lmur (82, 76), atlochur (49, 73), atamidiur (51, cf. 49), nomolur (75), domuiniur (58), fossuidiur (71) : active rotchlunim (72), laimim (57), rornolaim (71). The corresponding person in the subjunctive is regularly deponent ; we shall have the examples later. Pres. ind. sg. 2 ataigthw (66), adgladaither (54), admuinter (76)-atoZunisiu (66), acailh’ (69).l Pres. subj. sg. 22agithr (59), aigther (65), airlither (65), arafolmaider (1) (51)-cuirs (58, etc.), tochre (58). The 2 sg. ipv. is usualIy deponent cluinte (48, 53, 60, 75), cobairthe (70), cuirthe (58), jnta (61), at~nglaitese (48), moktha (75), atlaigths (47), samailte (56, 57, 87)-active cuir (66), accmail (57, 88). Pret. sg. 2 adrogailser (55). The third person is regularly deponent. s fut. and subj. condaesur (67), roessw, cgessar (55, cf. 61, etc.) ; rofeiser (f5), mesir (49), duthroiser (75); $astar (48, cf. 51, 72), admsstw (51), conmestar (67), rnidhthrastw (47) ; cofessamw (61), adglaasrnar (54), con- mesamw (54), condessamar (81) ; rofessatdr (57). Red. Fut. atagegallarsa (64) ; gignithw (70), adgignethar (61), atagegall- dathw (64), lthaither (61). The b fut. is scantily represented, cotnofadar (51), auriSfernmar (70)-immacurj (60), cotndba (51), nosseichfe (51). Our next task should properly have been to trace the decay of the deponent inflexion. But in the absence of a sufficiency of examples in dated and trustworthy texts we should here find ourselves on very slippery ground. The safer course is first to start from a time by which it can be shown that the deponent inflexion had disappeared, and then to work our way backwards as far as we can. Here we may first call attention to certain forms that survived for a while the general breaking-up of the deponentid system, and that are found in texts in which the deponent has otherwise passed into the active. Of these forms the 8 preterite and the

1 On atwiglaithe, see p. 69 n. a On forms like essara (52), nitdyara (65), see Zimmer, KZ. xxviii. 342 sq. In dogiglzestdr (58) we seem to have a mixture of the red. fut. with the s fut. Cf. the fut. pass. atatchigestar ‘r will be seen,’ MI. 59c 12, with dianaioiyther, LU. 74” 25, aeciyed, L . 64a 39. With the phrase dogignestcir do menma cf. niba sldnir de Int do menma, LU. 74b 24. THE DEPONENT VERB IN IRISH+. STRACHBN. 5.19

1 sg. subjunctive will be noticed in the third part of this paper. Apart from them, the chief instance of the kind is the -M perfect, which not only lived on but even extended itself analogically, cf. p. 93. Examples of this are atacaemnacair, conw- neew, etc. (SS), forcoimnacuir, rodamar, etc. (78), atamir, ginair (79), romidair, tathamair (80), rodcimair, lama&, dosfanair, diarcisair (81), coemnacair, rodamair, etc. (82), doruminair, etc. (84), fwcaomnmuir, cornmidair, etc. (85), rugenair, rolamair, etc. (87), niroddmar, etc. (89), roiarair (iairraim, LL. 158b 32), rocssair, conanacar, rodamair, forodamair, forcoemnaeaei; roghar, fwoer~angair, duthracair, tarrasair (Passions and Homilies from the Leabhar Breac, ed. Atkinson). If it be asked why this form outlived the rest of the deponent, the answer may be that it found support, which the other parts of the deponent did not, in forms outwardly similar from active verbs in which the r belonged to the verbal stem. Compare, for example, roginair with dorochair, dorochratar ‘ fell,’ rofrecair (for older frisrogart), rofrecratair answered,’ rolmmair (lommraim ‘ strip ’), etc. Even here the 8 preterite began at an early period to exert an influence. In the Wurzburg Glosses rumuinestar (18) seems to have replaced rwninaip; and in the Milan Glosses (ib.) romknatar has been ousted by the s preterite in fwuraithminset and mumuinset. Further examples of the same kind are narlam (65), rolamsatar (70)? rodamsatar (71), dorumensat (82), rofodaim (84), doruimm,‘ rolam, rodam (85)’ nirogein, ramidestar (87), roataim, rolam (88), romid rogenis, rolarn (89), nirlam (LU. 428 45), rogein (LL. 124b 4), rolam (LL. 271b 22). Of single verbs the deponent inflexion lived on in fetar ‘ know,’ which, in the phrase ni fhadar, has survived to the present day. In the later Hdiddle-Irish literary language the old forms are found transformed after the 8 pret., e.g. fedarais, fedrais, for 2 sg. fetar. The survival of the verb is probably due to the frequency of its use. Another deponent verb found in texts in which the active inflexion prevails is dgur ‘fear,’ cf. pp. 59, 85, 86, 87. The only instance of the active that I have noted is the ipv. aigsiu, LL. 278, lower margin. A remarkable thing about this verb is its complete disappearance

1 In the dmra ColuimBchil2e 108 dorunzeoin is doubtless to be explained w a momentarv analogical formation after -ge& ‘knew,’ cf. aadigdi gedin 66b. It is one of ‘the many artificial forms in that com osition, cf. p. 46 n. Another long poem composed in the same obscure and artilcial style is the Amra Conroi, H. 3. 18 (T.C.D.),49 sq., ascribed in the MS. to Ferohertne. Phil. Trans. 1891-2-3. 36 550 THE DEPONENT VERB IN IRISH-J. STRACHAN. in long texts like the Passions and Xorniliss, where it is replaced by immeclaagim as in late versions of the Fdin 3d Ctdailkge. 80 far as I can see, the only hypothesis that will explain the facts is that the verb vanished from the language in the deponent stage. In the later texts in which it occurs it is probably a purely literary survival. The absence of corresponding active forms would account for the retention of the verb as a deponent. In considering further the loss of the deponential inflexion we shall leave these forms and words out of account. The Zea6har na h-Uidhre was written before A.D. 1106, cf. Introduction to the Facsimile, p. xi. Hence any piece contained in it must belong at the latest to the eleventh century. Now side by side with the old texts in which the deponent inflexion is found, there is another series of texts in which it has disappeared, cf. p. 83 ; further, the Comthoth Loegairi (117b-l18b, cf. Zf.D.A. xxxv. 73), where stands fomatairisje 118a 27 by the s pret. &nntmusnarlasatar 117b 37, Cath Cairnd Conaill (115b- 1178), where we find nalaim 1168 3, 4, by atloohur 116b 26 (cf. p. 89), Imram Xaelduin, cf. p. 91. It follows from this that the deponent inflexion had been lost at the latest before the end of the eleventh century. The poems (p. 81) of Gilla Coemain t1072 and hie translation of Nennius (p. 89), the poems of Fland Mainistrech jl056 (p. Sl), and the poems of Cuan hua Lothchaiu t1024 (p. 80) carry us back towards the beginning of the eleventh century. In none of these are deponent forms found except in the parts of the verb referred to aboye, p. 105 sp. Mac Liac +lo15 (p. 80) has the deponent rofnnathar. It is quite in accordance with the usage of Irish literature that isolated instances of a formation should be found under the influence of literary tradition after the formation as a whole has been lost, and rojnnathar is a case in point. In the Borama, which may without hesitation be referred to this century, cf. p. 86 n., is found another example of the same kind, admoniur, which may be compared with atlochwr, p. 89. We may say with certainty that the deponential inflexion had, escept in special forms, disappeared by the beginning of the eleventh century. We come now to the tenth century. To the latter part of this century belong the Saltair na Rann, cf. p. 81 n., and the Tripartite Life of 8. Patrick, the latest reference in which is to a person who died A.D. 936, Stokes, Introduction, Ixiv., cf. Zimmer, Nmnius Pindieatus 210. We have also the poems ascribed to Cinaed hua THE DEPONENT VERB IN IRlSH-J. STRACHAN. 551

Artacain t975, and Dallitn Mac M6re, cf. p. 79 n. Here we may also mention the Aireo Menman Uraird Xaic Coisi (p. 85 n.), cf. Zimmer, Z$B.A. xxxv. 35. In the Saltair deponent forms are few. Of these condessama~~may have been used for the sake of the jingle with rofessamar; the words occur in two successive lines. The ipv. clzchte is found in somewhat late texts, e.g. LL. 269b 7, 272b 11. The form lamzcr is found also in the Trip. Life, and thus may be ascribed to at least the literary language of this century. Of folamadair in the Trip. Life I have already spoken, p. 93 n. In the verb cluinim the Saltair has ouly active forms; the two forms found in the Trip. Life are deponent. A couple of decades may have elapsed between the composition of the two works ; but whether the difference is to be explained in this way, or whether these deponent forms have been inherited from the sources from which the Trip. Life was compiled, I do not venture to decide. The latter explanation seems to me the more probable in the case of the isolated coimnigedar. The evidence seems to prove decisively that the deponent inflexion of verbs in -ig- had disappeared before the beginning of the tenth century. In the poems of Cinaed the isolated doomwemar is probably a literary survival. In the poems ascribed to Dalliin Mac M6e, which would thus fall about the beginning of the century, the examples are neither very numerous nor very decisive. The verb ouirim shows active forms in deponential texts, of. pp. 56, 66 ; similarly atlaigim, p. 74, if the reading atlaigit represents the original. The fut. cechla3 proves nothing for the present stem. As to curnnigim we have already seen that the deponent inflexion wae probably lost earlier in this class of verbs than in the other. On the facts before us we may, I think, say that the deponent inflexion had, except in certain forms, practically vanished by the middle of the tenth century, and we shall hardly be wrong in adding that in all probability it was at least well broken up by

1 The actiye maniessat occurs in the Battle of Mucrima, LL. 289b 22. Later esaaim becomes isaim, cf. p. 89, Atkinson, Passiolas and HOm&38, 8.v. ithim, Three 8hufta of Death, 8.v. ithim. The change of vowel may be ascribed to the influence of the pres. ithim ; ia- : ith- =ris- : rice-, etc. * It is interesting to note that similar form are found in the Siabu,Jwpat ConcuZaind (p. 68), which Zimmer dates from the middle of this century, and which, so far as I can judge now, need not be much earlier. 3 The same future conqechlat is found LL. 188c 13 in the Dialogue of the Two Sages, which, if O’Curry be right, dates from the tenth century, ef. p. 110. Cechla represents the fist stage after the dep. cechladar, and is older than cechlabat (p. 69) and atchlzlinfet (p. 84). 552 THE DEPONENT VERB IN IRISH-J. STRACHAN. the beginning of the century, though here additional evidence would be most welcome. As we try to penetrate yet further back, the ground becomes very shaky. There can be no reasonable doubt that at the beginning of the ninth century the deponent was still in a good state of preservation. But its history through the century I have been unable to trace with any degree of certainty. Possibly the poems ascribed, rightly or wrongly, to persons of this period might yield some results, though there great caution would have to be exercised, but I have been able to see but few of them, and these few have given little help. The various scraps preserved in the Four Masters contain under the year 845 nochongignethair 'will not be born,' under the year 848 atoluin, under the year 885 rommeBair by doruimen, but these dates fix only a superior limit for the compositions. A quatrain in LL. 278, ascribed by Atkinson to one Daniel t861, has (isur, a@&, robmur, laotchuirfe.' A poem, LL. 13@, containing adocorsat 134b 4, is said to have been written by Maelmuire Othna 1884, but the authorship has been rightly questioned by Zimmer, ZJD.A. xxxv. 83. As to the older prose texts, before anything very definite can be said, further investigation will be necessary both of the language and the matter, and, in the case of the heroic tales, it would be of the utmost moment to determine the precise amount of Teutonic influence on them, and the way by which it came. As matters stand, the wisest course seems to be to refrain from any too positive assertions about the history of the deponent during the ninth century, and to content ourselves with some general remarks on the texts that we have collected. First, as to the Glossary which passes under the name of Cormac Mac CuilennBin. In the beginning of my paper I referred to this work as a fixed landmark in the history of the deponent. Further reflexion has led me to modify this opinion. Before any certain inferences can be drawn from this composition with respect to the state of the deponential indexion at the end of the ninth century, it must be shown that the articles in which the deponent forms occur proceed from the pen of Cormac or some one of his time, and that they were written in the language in common use in the literature of the period. In the absence of any positive proof, the traditional authorship of the Glossary

1 Cf. the acthe immacurj, p. 60. THE DEPONENT VERB IN IRISH-J. STRACHAN. 553

could be accepted with confidence only if it could be shown that the language of the work is the language of the period to which it is ascribed. If that be so, in arguing from the Glossary as to the state of the deponent at the end of the ninth century, we should be moving in a vicious circle. Of Cormac’s connexion with the Glossary we know absolutely nothing. Even granting that he did put together some part of the work that bears his name, we do not know that he did not incorporate with it older material, the language of which, as the work was a learned one, he did not think it necessary to modernize; it can hardly have been first in the ninth century that the Irish showed themselves apt pupils of Isidore. It may be noted that the examples of the deponent come from a few articles, the longest of which, prulZ, is preserved as an independent text in Harleian 5280, British Museum, and in 23, N. 10, Royal Irish Academy. As to the language of these articles, it seems to me much more archaic than might have been expected in the end of the ninth century, though here it is difficult to get a certain standard of comparison. Note particularly the deponent inflexion in the -ig- verbs. I hinted at the possibility of a traditionally archaic language. Of this we seem to have an example in the Dialogue of the Two Sages with its asmidetar LL. 186b 41, docuridar 1868 24, 27, 29, 49, fuinethar 186b 43, if o’curry, on ths Xanusw+t Materials of Irish History, 385,’ be right in assigning it to the tenth century. We find even in the Commentary, which is obviously late, docuridar LL. 187a 52. But it is impossible to go into this question here. It will have, however, to be faced when the history of the Laws and of the treatises on grammar and rersification comes to be written. The answer to it does not affect the history of the regular development of the deponent in Irish. Another work which has been placed in the tenth century is the F6lire of Oengus. Chiefly from linguistic considerations Stokes, Introduction 11, concludes that the poem was written towards the end of the tenth century. Warren, fiturgy and Ritual of the Celtic Church, p. xv., says that it was certainly written after A.D. 982. But the state of the deponential inflexion seems to me to prove that the Calendar is at the latest not

1 The references in Cormae’s Glossary prove nothing to the contrary, for we do not know when they were put there, any more than the reference in the commentary on the Amra Colzkimbchille, p. 26. 554 THE DEPONENT VERB IN IRISH-J. STRACHAN. later than the ninth century. Compare it with the condition of affairs in the Saltair na Rann and other poems of the tenth century. And the language of the F6lire as a whole is, so far as I can judge, more archaic than the language of these tenth-century poems. Mr. Stokes admits that most of the Middle Irish corruptions may be due to the transcribers. Some of those that he would impute to the author seem to me to admit of the same explanation. Thus it is not impossible that the old neuter of the article was replaced by the masculine or feminine form, cf. similar corruptions p. 73n. In June 21 the verse hua liathain inligach would be improved by reading hua liathain Zigach; the insertion of in may be reckoned among the devices that later copyists had for restoring an apparently defective verse according to their lights. For hthair, trebthair, conicsid might be substituted without compunction the correct forms and who will guarantee that in Ep. 6 dosruirmisem has not been altered to dosrimemar, or in Prol. 25, etc., bebni to bebais? In March 11 the reading uayfaithib is better attested than itir Jlaithib; in November 23 there would be no difficulty in changing etir tonnaib to etir tonna, or in Ep. 315 in altering iarsinmbciigsi to ~ars~nbdigsi. Forms which admit of so easy correction can hardly be considered decisive proof of the lateness of the text. As to other forms which Stokes refers to, pp. 6, 7, it has to be proved that they cannot go back to, let us say, the beginning of the ninth century. For example, ma is found in the Reglum Mochuta, p. 73, and in other old texts. Mr. Warren’s assertion doubtless rests on the reference to the two Sinchells, March 26. But the Rawlinson and Laud copies mention only one Sinehell, and a poem quoted in the commentary in the Leabhar Breac refers to one only. The inference is that dasincheil of L.Br. is due to a corruption that got into that text after A.D. 982. As to the reference to Oengus himself, March 11, it seems clear from a comparison of the readings that Oengus did not stand in the original text. Further, if the Calendar had been composed in the tenth century, we might have expected to have had mention of some of the noted saints of the ninth and tenth centuries. I have had no time to hunt up all the saints in the Calendar, but O’Curry, Nanzcscr+t Xateriab 362, says that he knows of none later than A.D. 798. Again Donnchad t793 is referred to as the type of earthly glory. Why should he have been selected above all others in the tenth century? It may be that I have overlooked evidence from the

\ THE DEPONENT VERB IN IRISH-J. STRACHAN. 555

language or the matter, but until such evidence to the contrary is produced, I shall be content to refer the poem, whether composed by Oengus or not, to the ninth century. The so-called Reglum Nochuta Rathifa has already been discussed, p. 73. I should have preferred to give it an earlier place, but I did not make its acquaintance in Dublin till the fist seventy pages or so of my paper had been printed off. However, this matters the less as the arrangement of these old texts generally does not profess to be strictly chronological. As to the Irish Hymns I have nothing to add to what I have said already, except that most of them seem to me to be best placed in the ninth century. Section 5, pp. 72-3, does not call for special remark. It furnishes examples of the deponent in a special kind of texts of indeterminate date,. but which need not be later than the ninth century. We have still left sections 4 and 12, which form a considerable proportion of the first part of the paper, and which call for some further discussion. These tales may be divided roughly into two classes. In the one class the deponent inflexion is well preserved, in the other it has disappeared or survives only in fragments. This difference in language cannot, so far as I see, be explained except by difference in date of redaction. The deponential texts are more or less faithful copies of texts of the ninth century or earlier; the others are either of later origin altogether, or in them old texts have been remodelled in accordance with changes of language and circumstances. This is exactly what might have been expected. After all, the business of the aJZide, the professional story-teller, was to amuse, and he could hardly have held his audience enthralled by any tale, however exciting, which they, and possibly he, did not understand. Is the eridence of the language supported by the evidence of the subject-matter ? Zimmer has devoted a paper in the Zeitsch@ fur deutachs Alterthum xxxii. 196 EY. (cf. also xxxiii. 129 84. xxxv, 1 aq.) to the influence of the Teutons and particularly of the Norse

1 It would be interesting and it should be instructive to compare the language of these two sets of tales in other pointa. Much mav be learned from Prof. Zimmer’s paper, KZ. xxviii. 313 sq., which throws light on many thinga in Mid. Ir. grammar. There the examples of later formations come chiefly from latm texts, though sometimes they have crept in the course of transcription into the earlier ones. Here we must distinguish the different dates of origin of the Merent new formations. Thus, the 3 sg. pret. pass. in -ta, -tha is as old as the Irish Hymns, fechta ‘was fought’ ii. 67, and so may have stood from the outaet in texts of the ninth century. 556 THE DEPONENT VERB IN IRISH-J. STRACITAN.

on the literature and language of Ireland. As might have been expected, his clearest and most indisputable instances of Norse influence in the moulding of Irish tales come from our second cIass. See his remarks on the Aihd Quill (85), p. 208sp., on the Cath -Ruis8 na Rig (88), p. 219 sp., on the LL. version of the dlesca Ulad (M),p. 265, note. Of the relation between the LU. and LL. versions of the Tdin I have already said something (58 note). Here again the clear traces of Norse influence are found in the LL. version, ib. pp. 204 sq., 263 84. In the passage, LU. 79 q., which agrees verbally with LL., and which, as I have remarked, p. 59, looks as though it were a latter addition, Zimmer seems to me to be right in seeing Norse influence, ib. 3178p.l In one of the tales which, because of certain deponential forms I put in the first class, but which it would be better to place on the border line between the two, the Siaburcharpat Conculaind (68), Norse influence is clear enough, Zimmer, ib. 248 sp. As we have seen, this piece may be put in the first part of the tenth century, when the deponent was - drawing its last breath. So in the Tochmarc Emero (71), Zimmer, ib. 239, may be right in seeing Norse influence in the episode, LU. 126a 11-41. He is followed by Yeyer; Rev. Celt. xi. 438, who sees in this version of the tale a post-Norse redaction. On the part of the deponent there seem to be no insuperable diBculties in the way of this. The only two decisive forms, adg[l]ddur and fosszcidiur, are found in a passage very like one in the Compert Colzculaind, LU. 128b, and which may have been some sort of traditional formula, on which the modernizer did not dare to lay his hands. But as to this I should not care to express a decided opinion from the short fragments in LU.

In part of the LU. version Zimmer, ih. 308 sp., sees a more subtle influence of the story of the Nibelungs. Unfortunately I am too little acquainted with the Teutonic tale in its various forms to follow all the reasoning, but the resemblance seem to me to be too slight to draw any conclusive inferences from. And could the story of the Nibelungs have reached Ireland only through Norway? Against Zimmer’s hypothesis is the language. I see no evidence that this passage is later in language than the rest of the LU. version. And unless ow conclusions with regard to the disa pearance of the deponent are altogether wron that version cannot at the Patest be later than the ninth century, and I sfodd be inclined to say not later than about the middle of it. And such far-reaching Norse influence would at that period be out of the question. a He has published, Rev. Celt. xi. 442 aq., what he considers a pre-Norse version. Here are found deponent adgladastar 1. 23, cf. 11. 47, 63, roairigestar 1. 42, and analogically dzlfohtrastar 1. 98 ; active-cotnacalt 1. 10, molsi 1. 64. THE DEPONENT VERB IN IRISH-J. STRACHAN. 557

As to the other tales which I have included in this class, Zimmer, ib. 332, sees Norse influence in the Fled Bricrend (p. SS), in the fight on horseback with Ercoil. If I mistake not, in a paper in the Sitsber. der Berliner Akademie, he has shown that the Irish knew of the Norse form of sport, the hestavig, before the Vikings established themselves in Ireland. As to a direct reference to Norsemen in these texts, that depends, so far as I see, on Zimmer’s interpretation of jann and dibergach. As is well known, Zimmer, z;f.B.A. xxxv. 12 sq., identifiesJiann with the NorsejandR ‘enemy,’ and regards it as the first name by which the Irish knew these ruthless invaders. If that were so, it would bring Norsemen into many of these early texts, such as the Fled Bricrend (p. 55), the Orgain Bruidne Da derga, the Xckl Xuicce Naic Ddthd,‘ and the LU. version of the Tdin.‘ In many ways the theory is attractive, but the objections against it seem to me to be much stronger than the arguments in its favour. As to dibergach, his identification of it with an imaginary Norse Tyverk (Gott. gel. Ans. 1891, p. 195) is yet more improbable. At all events the word can be traced to a considerable antiquity. In G10.sm Hibernim 284 Zimmer corrects iddemergach very probably to aithdibergach, cf. introduction, p. dv. In the Arrada, Rawlinson 512, B. 42b 2, diberg is mentioned along with many other sins-shut rongabsat jhagala 7 duineorcni 7 duinetciidi 7 sicut rogakat diberga 7 druidechta, etc., ‘such as are fratricides and homicides and secret murders with concealment of the body, and such as are diberga and sorceries,’ etc.2 It is d priori improbable that the

1 Zimmer geema to suppose that the word may have got into the Irish Sagas in the first part of the ninth century. 2 In ZJ0.A. xxe. 267 sq., 464 8q., cf. xxxv. 19, etc. (cf. also Meyer, Rev. Celt. xi. 493 sq., xii. 460 aq., Stokes BB. xviii. 122 sq.), Zimmer notes a number of Teutonic loan-words in Irish. As might have been expected, these words abound in the later texts. Here loan-words from Norse may be looked for in large numbers. In the old texts supposed Norse loan-words must be regarded with suspicion and strictly questioned as to their origin. Here the probabilities are that any Teutonic words came rather from Anglo-Saxon. Zimmer’s identification (p. 466) of the common word dZ ‘drink’ with Norse OZ can be proved wrong. The word is already found in the Milan Glosses 94e 12, 118C 9, 129d 14. Moreover, 61 was originally disyllabic, as is shown by the line in the story of Bran Mac Febail, Rawlinson 512 B. 119a 2, hool fino oengrinde. I have also seen it disyllabic in the SuZtuir na Rann, but I cannot find the reference. Zimmer (pp. 279 sq.) derives Zuith ‘beer’ from a purely hypothetical Teutonic Zaipu- or Zuipo-, hut a derivative (in) laithertach g. crapulatus is found M1. 1OOa 10, and Stokes UrJceZt. Spruchschatz compares lacth with Corn. lad g. liquor. In other instances Teutonic origin is uncertain or improbable. There is no strong proof that rudiud (p. 278) is borrowed. If I am right in deriving rucce ‘disgrace’ from *rudcili the grade rud- is also Keltic, 558 THE DEPOKENT VERB IN IRISH-$. STRACHAN.

Vikings should figure in tales of so early a date, and much more conclusive evidence will be required before their presence can be accepted as an established fact. It does not belong to the subject of this paper to enquire minutely into the date of redaction of the old heroic tales. At the most the deponent inflexion can only help to fix an inferior limit. It is to be regretted that we should have been unable to find surer data for the history of the deponent in the ninth century. Judging simply from probabilities I should hesitate to place any tale in which the deponent inflexion is well preserved later, at the latest, than about the middle of the ninth century, but that as yet is only a subjective opinion. These tales need not all belong to the same period. Thus, the simple stones Ydin Bd Regmain and Tdin B6 Regamna furnish, if my observations are accurate, no examples of preterites like ga6ai.s which are found in other tales. But an answer to such questions could be found only by a detailed investigation and comparison of the language of the several texts.’

111. NEW EXTENSIONS OF THE DEPONENT.

1. THE a PRETERITE. Deponential forms spread by analogy in the s preterite, first in the 3 sg.,2 later in the 1 pl. and 3 pl. In the Old Irish Glosses there are only a very few instances, all in the 3 ~g.~:-rodligestar,

The derivation of fuine ‘cooking,’ fonoim ‘cook’ from Norse fun; ‘fire’ (xxxv. 159 n.), is altogether improbable. Nor is it at all certain that gilla is borrowed from Norse giZdR (xxxv. 15). 1 A few words may be not amiss with regard to one of our texts, the Serglige Cbnczllaind (p. 65). Here it is to be noted that most of the decisive deponent forms come from the Briatharthecosc Connclaind, which reminds one of the Tecosca Cmaic, LL. 343. The language of this seems to be more archaic than the language of the rest of the text, in which, for example, the pres. in -9zd is found repeatedly even in Verne, funend, laicharand, Izinaceend. p Probably because the need was felt of a more distinctive ending. 3 I know of no such analogical formation in the 1 sg. Hence forderisircq-, M1 133b 8, must stand on a different footing. According to Ascoli it gloses lustravi. The whole sentence runs--excelsa quceque aspectec Zustravi suqerans expectana solutionem malorum. Perhaps forderissiur is to be analysed into for-&- ro-sissiur, a compound of which I have no further examples. As to forhet, Ml. 78b 8, it can have no comexion. Doesfordere! come fromfor-di-vethid? THE DEPONENT VERR IN IRISH-J. BTRACHAN. 559 arrzcneastar, a&oneestar, arruneillestm (25). In the other deponential texts the formation is either not found at all or is infrequent, and it is almost confined to the 3 sg. :--sknastar, rodglinnestar, noconmellestar,' aridralastar (47), diuchrastair (55), romarastar (56), rodligestcir, immcirdttnestdr (57), rocarastar, condacermnastar, dorkt- laa'stir, rogabastclr (63), rocharastar, coromaichnestnr (SS), roders- cat$estar, arlastdr, tallastdr,faatbestdr (68), beogastar (69), bdlgestar (71), roscarastar (72), rotordnestar (73). Examples of the 3 pl. are rare, and come from texts that show a certain breaking up of the deponential inflexion :-ro$ersatdr (57), roscarsatar (66), ochaatar (68), cotairchdlsatar (73). The 1 pl. is found in roinnisemw (63, in a parenthetical remark of the writer), rophenad- 8emmar (72). Let us aee now how the forms are distributed in the later texts.' In the Tripartite Life the examples are still almost all active (78 n.). In the secular poems ascribed to writers of the tenth and eleventh centuries the 3 sg. only is found in the poems of DallBn Mac More (79) and Fland Mainistrech (81). In poems ascribed to Eochaid hua Flaind, Cuan hua Lothchain, and Cfilla Coemain (80, 81) are found pl. forms; the examples are found chiefly in a couple of poems attributed to the second writer. In the Saltair na Rann examples are much more numerous in the plural than in the singular. In some of the later tales the singular prevails, in the Akec Xenmala Uraird illaic Coisi (probably end of tenth century), and in the Deaths of Goll and Garb; in others the pl., as in Cath Rwiss na Rig (where the sg. and pl. are pretty evenly balanced), the Horoma, and the later portion of the LL. version of the Tdin. This would seem to indicate a difference in the time of composition. The earliest of our texts which show numerous plural forms is the Saltair na Rann, in the end of the tenth century. But the use of the one form or the other in the plural would seem to have been, to some extent at least and during a certain period, a matter of individual taste, for, as we have seen, Fland

1 The three first instances all come from one hymn. 2 In these texts we have also forms from the 8 preterite of -up, -is- verbs, which in O.Ir. were prevalently deponent, and in the 3 sg. 8. pret. always so. Without going into detail, we may note here how in the 3 sg. the deponential inflexion tends to dieappear, while it shows a tendency to spread again in the 3 pl. Compare the instances in the Tripartite Life with those in the 8aZtaw na Ram, the Togail noi, and the Cath &Ass lza Rig. 560 THE DEPONENT VERB TN IRISH--J. STRACBAN.

Mainistrech t1056 regularly uses the singular. For the later history of these forms I have no collections of material.'

2. THE 1 SINGULARSwmcr~w.

Here the deponential inflexion spread by analogy, rue, for example, becoming rucur, dognko, dognCor. The reason for this may be found in the convenience of having a more distinctive ending for this person than was supplied by the active forms, such as fadam Wb. 29d 27, pridaeh 10d 25, tormal MI. 109b 6, eonrzil-e'lsa 4Id 12. Owing to the paucity of the examples and the suspicion that later forms were introduced into old texts, it is impossible to give a detailed history-of the formation, and to trace with certainty the various stages of its development. I shall content myself with putting together the instances I have noted in the above textsl2 and then suggesting some possibilities as to the spread of the form. In the Glosses the deponent ending is found only in deponent 1-erbs, including the verbs in -ag-, -ig-, which in this person are regularly deponent :- labrar, -samlar, frisaccar (orthotonic adcear, p. 23 n.), colhuthlucher, erladaigear,3 nufailtiger, noailaiger, cloor, -mesur. In the other texts the treatment of the present subjunctive differs from that of the subjunctive of the a aorist, and it will be better to take the two tenses separately. First as to the present- TQinB6 FriLich. Active forms :-condatucsa 2518 34, niconb6oosa 25 la 38. Ec6l Muicce Maic DQth6. Deponent form from deponent verb :- corotacilliur 112b 47. Longes Mac n-Usnig. Deponent =old deponent :-conidnaecur 2608 24.

1 Further examples may be found in Atkinson's Glossary to his Pasaims and €fomiZieu. I might just call attention to the great fondness for the deponent forms in the glosses on the Irish Hymns. 2 sp. Further instances from LU. and LL. mill be found in EZ. xxvii. 347 . 3 Cf. p. 34 n. If -gear is the reading of the MS., it is probably due to the hesitation of the scribe between -er and -ar ; or could -era- have been suggested by the active ending -ea in the third conjugation, e.g. cofotheaea, M1. 38 9 P THE DEPONENT VERB IN IRISH-J. STRACHAN. 561

Togail Bruidne Da Derga. Deponent forms from active verbs :-con&& 83b 14, conidra& 84b 32. Tdin B6 CuaiLge, LU. Active forms :-nidko 73b 6, ceinnonbko 746 36, cofidatuc 15, 5gb 6, oowucsu 69& 5, cofarcaka 69a 5, conidndkrtia 62b 13. Deponent forms from active verbs :-conapror 828 26, aondwnar 708 20, oonidrokrsa 63b 15, cotopaahtur 73b 2. Fled Bricrend. Deponent forms from active verbs :-corochotlur 1128 47, manqetur 100a 8, cotallursa 104& 15. Siaburcharpat Conculaind. Deponent forms from deponent verbs :-conidnarladzcr 1138 7, conidnacur 1138 6. Short stories, etc. (p. 72). Active form :-eonerhorscs 2798 25. Deponent forms from deponent verbs :-c& 2808 19, conidnaccur 282& 1. Deponent form from active verb :-demur 2796 30. FQlire of Oengus. Active form :--robeo,Prol. 271, 273, Ep. 315. Deponent form from active verb :-possibly ronruidiur, Prol. 277. Saltair na Rann. Active forms :-be0 1844, 1540, nadevn 1563, coiromarb 5828, coruc~a1595, conostuc 1665, eotuc 5827, dianamthacr- biur if I humble myself,' 844. Deponent forms from active verbs :-corobadur 1667, arnahhrbalw 1260, contorchrwsa 1533, cona8rucur 1666. The remaining texts, except in one case, show only forms in r :- Trip. Life :-eonacor 52. 24, coroehreitizcr 46. 23, corothadcuirer 180. 11, corofoihigiur 52. 22, niunitoindiur 200. 13, )Jut ceB mbkosa 116. 20. Cum hua Lothchain :-cofargbur, LL. 209" 27.

1 At 71s 14 nadbmuv is, as Zimmer has pointed out, a glosa on nadneoma, which has crept into the text. 562 THE DEPONENT VERB IN IRISH-J. STRACHAN.

Fland Nainistrech :-corinniseor, LL. lla 19. Gilla Coemain :-eoro~nnis~w,LL. 12Ta 37. Airec Menman Uraid Maic coisi :-conerbharsa. T6n B6 CuaiGge, LL. :-rnenibenur 102b 34, cancorabur 6? 7, gqfaccursa 888 21, doneor 81b 26, condernur 81b 28, dh- nomgluaisiur 67b 14, danamluur 67b 17, corucur 728 31, 26, conastardur 638 10, corotholiur 82b 30, from reduplicated form' gmosiblur 1038 43, cf. 1. 46. Deaths of Go11 and Garb :-anerE beorsa 108b 32, nada~accursa 1088 7, cofagbur I08b 24, corofrornwr 110b 35. Borama :-cwoche?anagur 300b 39, rnanidiglmsa 296b 37, but ckilzbeo 3078 3.2 Cath Ruiss na Rig :-corochuriw 175%39, condigiwrsa3 1738 39. I have not noted the instances in the Togail Troi. We see that the 1 sg. subj. of the old deponent verbs still remains deponent, only with a change of vowel of which we shall have to speak later. I see no reason for supposing that an active stage came between the old forms in -ar, -w and the later forms in -ur. As to the date of the substitution of -ur and the spread of -2cr to verbs originally active, it is impossible to speak with certainty from the evidence before us. In any texts -w might very easily have been altered by copyists to the familitir -UP, and in prose texts forms like rue2cr may have been substituted for forms like rue. On the other hand, I have found no evidence that forms in -ur were unknown, let us sap, in the ninth century. In the FQlire conruidiup would be an example of the kind, if my suggestion about it be right, but it is not absolutely certain. It would be simpler, as we have seen (p. 98), to take asmecnugur in MI. as a subj., but it is possible that the glossator misunderstood his Latin. The Saltair na ltann sho\rs that the shorter forms still lived, in literature at all events, towards the end of the tenth century. The question whether they may not, for metrical convenience, have been kept longer

1 The future stem. 2 In maniimriu 296b 37 we seem to have an active form, if the reading of the facsimile be right. 3 This form has been put here because it contains no 8. But condigiur, like connechur LL. 1158 36, is based on an active form of the s subjunctive of dochad, cf. emuieeha in Glossary to Passsm a& Hoinilies 632a, and, for the vocalism of digiur, condiyussa Windisch, Wb. 468b. THE DEPONENT VERB IN IRISH-J. STRACHAN. 563 in verse is one which does not readily admit of an answer. The other poems of about the same time show one form, and that a deponent fmgbw, and that is too slight a basis to argue from. At aU events, from the beginning of the eleventh centq the deponent form was the rule. There is one notable exception, the subj. 660. Add to the examples given above hiceinbko LU. 408 41, din nombeosa LL. 26gb 3. The only explanation that I can suggest for this is that bko may have become associated with the adj. 640 in expressions like nochor'beosa LL. 2728 52, inhed bam 660, din bes €do, and the like. Even here a couple of examples of bkor are found, cf. Deaths of Goll, etc., above, and ckin korsa bko LL. 2038 45. The disappearance of this form is closely connected with the decay of the subjunctive mood generally, and falls in a period for which I have no collections. It is still found sometimes in Keating, probably as an archaism. I have noted a few cases in which it is already syntactically superseded by the secondary present, LL. 176b 18, incath barthoesum . . . . corocherkind, similarly 11. 26, 31, 37 ; LL. 1088 12, fodail innechraid . . . arbith colztulind; further, Battle of Ventry, 11. 175, 433, 625, 687, 960. In all these instances the Old Irish rules for the sequence of tenses would have required the subjunctive. The subjunctive of the s aorist has fared differently. In our texts deponent forms appear from old deponent verbs, conaaesur (67), cofessur (61), cofiasur (65), cofessur (82), cofesser (85). And, apparently from the similarity of the sound of the words, the deponent ending was extended from fiasur, fessurl to th from tlagaim, and tis from do-iccim, e.g. tiasur Orgain Bruidne Da Derga 97b 7, gotiasur Thin, LL. 1028 42, cotisor Thin, LU. 668 6 (but cotisa, 1. S), naaniththisiw Aided Conculaind 121b 10. Otherwise active forms, corisa, LU. 58" 20, corrius 62b 21, conkciw, colzdeochus 708 19, condechos 1298 10, conathoithu8, LL. 328 34, condkcius 200s 13, corius, BB. 462b 43.

1 This particular verb, as we have seen, lived long 88 a deponent. According fo Mr. Stokes, rofesseris the only old deponent form found in Gorman'a Martyrology, which he says waa written between 1160 and 1174. Under the influence of fiastar, etc., a dep. future is formed from adfhdaaim 'relate,' cf. 85 n., ciatfear Fled Bricrend 7 Loinges Mac n-Duil Dermait 1. 291, adfesar Trip. Life 222 1. 5; act. adfessam LL. llb 48, 131b 35. A further coincidence with forms of &tap. is found in the pret. peSS. adfessa Lu. 59. 75adfeta LL. 628 21. At U.123b 35 rojhzlr seems either corrupt or a mkformation. 504 THE DEPONENT VERB IN IRISH-J. STRACHAN.

In the Glosses -ur 1 appears in the middle subjunctive of the 8 aorist, -mesur? This is as might have been expected.' 10 the active a present subj. like dober, stands to an 8 aor. subj. like -pius in the aame relation as Lat. &am to dix6; the former ended originally in -CW, the latter in -6. To -6 in the active would correspond, as in the pres. ind., -6r in the middle, -mew-: corrius=lamur : do-6iur. Thus, mesur, fesaw, eaur would be regular developments. In the later language many of the pres. subjunctives in -ur could be derived from the old deponent subjunctive by the change of -w,-er, to -ur, e,g. -atcar, -accur ; Sarladar, arladur ; failtiger, cennagur. It is hard to say whether we have here simply the influence of -ur of the 8 aorist along with that of door, or whether there may not have been also some confusion of the subjunctive with the indicative forms, a confusion that would have been helped by their coincidence in the 8 forms. As in the deponent -ar3 became -ur, so in the active berim, erbar became erbor, erbur.4 Dogdo became by the simple addition of r dognCw, similarly bCo when it underwent any change. Derna became dernar and that,

On p. 6 n. I suggested that door is an 8 suhjunctive=*cZeu8Zr. But against this is the quantity of the vowel, for, so far as I have observed, the first o is short except in Eontracted forms like cZ6r. Cloor might come from *cZus6r, but we should rather expect the grade clew-. So it seems best to take it from a presential cleu-. The ending -or may seem strange in a present, but cf. bio to biu, eoniiZu MI. 86 '4 to glenina, dogne'o to dopniu. One thing common to all these forms is the loss of an internal consonant; so far as I know, the vocalism has not yet been explained. In the same way the 1 sg. subj. of aithgninim would probably be aithgnlu; the 3 sg. aithpne' is found LU. 718 34, cf. aithgnead LU. 72a 25. 2 On p. 16 dmmessursa has by a slip been printed for dummessarsa. The a is strange. Aseoli corrects to dommessursa, the regular dep. form. Can the glossator have imagined he was translating a passive ? 3 An ending -ar is found in conidnarlasar LU. 40a 31, corolamnar LL. 1268 9. Considering that these texts are apparently later than others containing -or, -ur, it is unsafe to connect -ar here with the old deponential -ar. It must rather be regarded as a new analogical formation, possibly after dernar and the like. In later manuscripts -at in this person stands on a different footing; there it may he explained from the confusion of vowels in final syllables. 4 Zimmer, KZ. xxviii. 348, ascribes the spread of the deponential form to the influence of the compounds of the root ber (mbiur, dobiur, forbiur, etc.)- " da bei hendie erste sing. activi des conjunctivs auf or, ur ausging (conerbor, diarzepw, manithabur, etc.) "-but surely the regular ending would have been -her, -bar, and the o is due to analogy. It is possible that besides the deponent yerbs the subjunctive -cw (jristacor MI. 118c 3) may have helped in bringing about the change to 0. 5 Similarly arld becomes arldr, LL. 204b 49. THE DEPONENT VERB IN IRISH-J. STRACHAN. 565 under the influence of the usual -ur, dernur.' In .other verbs the ending -ur was added bodily, rucur, benur, tardur, etc. SO from berim -bor became &or, and we even find for fessur, fessarur (Atkinson, Passions, etc., S.V. fetar), seemingly a new formation to 2 sg. fessara. In deponents of the third conjugation the ending -er may be postulated, %ider, etc., cf. duthlucher and examples from verbs in -giur. In later texts the usual ending here is 4ur, rnidiurla cuiriur (in Glosses active fristacor), gluaisiw, senniur, innisiur, etc.5 The reamn for this form can only be conjectured. It may be noted that both in the dep. "rnidw and the act. Zeicea the preceding consonant was slender, as indeed it was throughout this class of verbs. Hence it may very well have been that the slender quality of the consonant was preserved in the remodelled 1 sg. subj. and was expressed graphically by the following i. It is probably only a coincidence that the new subj. rnidiur is the same as the old indicative. But before anything can be asserted with certainty, it will be necessary to have examples from good copies of old texts. This ending -iur seems to have extended further in tisiur, condigiur, cf. also tail-biur, Saltair na Ram, 844. In Mod. Ir. the 2 sg. ends in -ir. The full development of this ending falls long after the period to which onr texts belong. But it may not be without interest to note the few forms that I have met, which indicate that this was originally a subjunctive ending, formed analogically to the 1 sg. in -ur. The only examples that I: have noted from LL. are arnatiasair 658 26, and acht natisirsiu 262b 28, but act. gotisiu 102a 53. These are evidently new formations to tiasur, tisor, perhaps under the influence of fesser : fessur, -bir : -biur, -bur. Similarly cornairser, Laws iv. 18. In the Glossary to the Passions and Homilies I have noted two present subjunctives, diijklngir and corolelzair, cf. @&air ' look,' 1. 7543 ; in 1. 7541 dingir may be subjunctive.

Similarly from rola ' went ' is formed a subj. rolur. In LU. 83b 14 comes the phrase condrldr teinid. Is drldr here formed on *drZi, *adroshi, an s subi. from ad-slaidim, cf. W. lladd tan? In 84b 32 -raldr is formed from -ra& Zaaim 'throw.' a Presupposed by -mider, Passions and Homilies. The -er here has nothing to do with the old deponent ending -w,but is a phonetic development of &r, cf. in the same work ronailes by rosailiws, dorinnes by dorignius, etc. But in LL. 11' 19 inniseor is clearly modelled on the active *innisea. Phil. Trans. 1891-2-5. 3i 566 THE DEPONENT VERB IN IRISHA. STRACHAN.

But the old form is still the common one. From the Book of Ballymote I have noted a couple of indicative forms, soich$uma (by yeha, dobha) 462b 31, fogebairsiu 481b 2. In the Acallam na Sen6rach I have noted but few forms, subjunctive ddtucnirsi 114, 1. 34, coJinnair 107, 1. 5, fut. ind. dogthair 225, 1. 7. Active forms are numerous. But to pursue the subject further lies beyond the scope of this paper, which is already long enough.