LETTERS FROM K. BŪGA TO J. ENDZELYNUi

Prepared by J. Karaciejus and A. Sabaliauskas

Among K. Būga’s epistolary materials, his letters to the Latvian linguist Jānis Endzelīns (1873–1961) occupy a particularly important place. The correspondence between these two foremost scholars of the Baltic languages began in 1909 and continued until K. Būga’s death. The correspondence is especially valuable because it deals almost entirely with problems in Baltic linguistics. Understandably, it also contains interesting observations that help clarify various biographical facts about both scholars. Most of K. Būga’s correspondence to J. Endzelīns is held in the manuscript collection of the Institute of the Lithuanian Language and Literature. Eleven letters held there were published in „Literatūroje ir kalboje“ (V., 1956, vol. 1, pp. 364–386; prepared by E. Grinaveckienė and J. Kruopas). Z. Zinkevičius’s edition of K. Būga’s „Rinktiniai raštai“ (V., 1961, vol. 3, pp. 817–879) includes 44 letters (36 from the manuscript collection of the Institute of the Lithuanian Language and Literature, including 10 previously published in „Literatūroje ir kalboje“, and 8 from the manuscript collection of the National Library). Z. Zinkevičius also gives the contents of unpublished letters known to him (vol. 3, pp. 994–996). Discussing K. Būga’s letters to J. Endzelīns, Z. Zinkevičius notes that „there are also some with the addressee“ (vol. 3, p. 817), but provides no description of these letters.

In February 1979, the Latvian linguist M. Saulė-Sleinė, J. Endzelīns’s life partner, sent the author of these lines 11 letters from K. Būga, written between 1911 and 1921, together with a covering letter (I.31). One of these letters (1921.X.9), based on the surviving draft, was published in K. Būga’s „Rinktiniai raštai“ (vol. 3, pp. 872–873). The draft and original texts differ in a few linguistic and stylistic revisions (for example, where the draft has susivažiavimo, the original uses susirinkimas; where it has žadu duoti trumpą kritiką, it reads žadu parašyti trumpą kritiką, and so on).

Along with the letters mentioned above, M. Saulė-Sleinė sent copies, handwritten by K. Būga himself, of his letters to the German linguist M. Fasmer (Vasmer, 1886–1962) and the Finnish archaeologist A. M. Tallgren (Tallgren, 1885–1945), who worked at the University of Tartu from 1920 to 1923.

Laiškas M. Fasmeriui (1921.IV.12) pagal Vilniaus universiteto bibliotekos rankraštyne saugomą laiško juodraštį yra paskelbtas K. Būgos „Rinktiniuose raštuose“ (t. 3, p. 872 – 873). Tačiau šiuose tekstuose yra ir skirtumų.

Among the materials sent was also one long letter (15 pages), without a date or signature. In this document, K. Būga appears to be answering certain questions. The answers are numbered (26).

The epistolary materials by K. Būga sent by M. Saulė-Sleinė were transferred to the manuscript collection of the Institute of the Lithuanian Language and Literature.

A. Sabaliauskas

54

[1]

Vvedenskaya 10, apt. 76.

St Petersburg, January 5, 1911.

Dear Friend!

Happy New Year!

I received the offprint from KZ. Thank you! I did not write immediately because I truly did not know where you would be during the holidays. A few observations on your article “О судьбе носовых звуков”...: 1) Low German sundau 3 = siundau Kvedarna, 2) Lithuanian stamančioti 7 is comparable to stamantrūs ‘springy’ Šiauliai (= Шавли, cf. Mież. 227 ~ stamantus without r!), 3) ruòbti 8 may also be of Lithuanian origin: iš-ruòpti, raupýti Jušk. dictionary I 622, 4) Latvian kampis 9: how should Latvian kanapis be understood? 5) Latvian spēdz 15 may also correspond to Lithuanian spiegia, spiègti ‘to cry out in a thin, high voice’ Dusėtos, 6) jėkti 21 from apj-ėkti (with e from a after j) ~ api+ąkti: a is indicated by the Kupiškėnai form apjǵaᵏᵒᵃ (õ from ã, cf. mõatᵒᵃ = mãto), 7) vagúoti 22 is unknown; only vagóti is found (Kvedarna, Kretinga, Šiauliai, Dusėtos, Linkmenės, Seinai), 8) glúosnis 23: glúosis is also known (without n!) Jūžintai.

Regarding KZ.: 9) Latvian trękns 57 = Lithuanian trėknas ‘fat’ Jušk. dictionary I 314: traknùs ‘harvest, abundant?’ = ‘luxuriant’ loc. cit. 4, 10) būžiej ‘insects’ 64 = būžiai ‘insects’, sg. būžys, būžio Butrimónys (plural), Trakai district, Vilna governorate, 11) aìnpalas ‘ice crust’ 65 from aĩ[t]-palas = iš-palas “the same” Dusėtos ~ pìlti: these words should be distinguished from Latvian pal’i ‘marshy lakeshore’ Biel. LŠpr. I 54 ~ pàlios, pàlių ‘large expanses of bogs = and marshes (Dusėtos) = kūdros (Vilniaus Kaltnėnai)’ Pilviškiai, Suvalkai governorate ~ Latin palūs, 12) saùbti, sautėti, apsaùsti 66 and following are known from dialects where š has shifted to s; I know only siaùbti (Low Lithuanian), siaùsti (High and Low Lithuanian), 13) kakù brėkumus 69: šaukinas ‘verst’ Kvedarna (Daukša’s Post.).

Daukša’s language is a very precious source for the Lithuanian language: it is full of material for every purpose! Daukša’s accentuation is quite distinctive. Instead of ugnis, šaknis, mirtis etc., Daukša’s language had ùgnis, šàknis, mìrtis!! There is nothing strange about that! The same apparent “misplacement” of the accent is also found in present-day dialects. Ūgnis feminine, grąžtis feminine, etc., are found in Dusėtos, Salos, Kamajai, and Prienai.

Kà man is known to me only from the expression — kà výrai, tà výrai! Kvedarna “what men, such men” (Dusėtos)! = молодцы!” cf. kàs tà kàìtris Jušk. dictionary I 421 b “what heat!”

Pri-ie-liešti žmonės DP. 364₂₂ “pri-liestė-ji”: liečiù, liẽsti “to oppress” ~ liecę ‘they press’ DP. 136₃₁, liecże 136₃₄, liecze 363₇, ~ pa-lytėti: 3rd person present pá-liti DP. 254₁₇. + liète naPlas “they oppressed widows” = spáudė našlės” DP. 24₉: liečiù, liẽsti “to touch, move” Dusėtos ~ negali jō až-lièst kaip vāpsvų gūžtōs! Ds. “not to be able to touch him... like a wasps’ nest.”

Jurkšaitis’s jūs is a genuine form, also known elsewhere (in Jablonskis’s dialect: Naūmiestis = Владиславов, Suvalkai governorate). Cf. Jablonskis (pseud. P. Kriaušaitis), Lietuviškos kalbos gramatika. Tilsit, 1901, p. 23: jūs.

Paskui could also be locative singular Pàsku (alongside Low Lithuanian paskuõ and pãskuo); this would more likely indicate that it is dative singular; cf. dovanai duoti ‘to give as a gift’ (the adjective *dóvanas is not attested).

55

Brugmann’s acc. pl. vilkuns is taken from A. Leskien’s Declination 104, where we read: msc. a — st. lit. vilkùs = vilkuns, aus *vilka-ns, -ans noch preussisch. We find *vilkuns neither in ancient texts nor in present-day speech. Such a form is theoretically impossible, too. -uns- acc. pl., -unse loc. pl. are known only in the Samogitian dialect; Highland Lithuanians have not heard them. Could this perhaps be a new nasalization, similar to Samogitian mąstis = Highland mostis < Russian масть ‘мазь’, mėš „мбы“ (Kalnujai near Raseiniai) = Highland mėš (Dusetos, Naūmiestis), glūnsnis (in L. Iwinskis’s calendars) = glūsnis Kved. ~ glōusnis Rietavas = Highland glūosnis?

Janiaus’s grammar has already finished printing; it will be sold only when the Russian translation, which I have been commissioned to prepare, is published. I have promised to write a spelling (orthographia) handbook. I think I will finish that work in the summer. Jablonskis began printing his syntax a month ago in the newspaper “Vadovas” (published in Sejny); it will be reprinted as a separate book. The etymology is being written by Jurgis Šlapelis (Jablonskis’s student) and will probably be printed by autumn.

Šlapelis is taking over Juškevičius’s dictionary from the letter m; so far he has printed only three gatherings. The letters k–l of Jušk.’s dictionary are with Ulijanov. Ulijanov has done nothing so far, and it is hard to hope that he will do anything. If it is difficult for a Lithuanian who knows his own language well to edit Jušk.’s dictionary, what can one say of Ulijanov, who has already largely forgotten it? We will probably not see Juškevičius’s dictionary completed for a long time! That would be a pity. The science of languages could gain very, very much from Juškevičius!

I cannot at present fulfill your request to review the literature on dat. pl. and loc. sg. I am very short of time! I could do it at the end of January.

While in Vilnius, you should first of all make the acquaintance of Dr. Jonas Basanavičius, chairman of the Science Society (if you have time, you could make much use of the library and archive), and Antanas Smetona, editor of “Viltis” (he has a fairly good practical command of the language). “Viltis”’s address: Кожевенная 5, кв. 25. You can find out from Basanavičius or Smetona whom else it would be worth meeting. It would be worth meeting the activist, Rev. Tumas, who until recently was editor of “Viltis” but, thanks to the Polish Endeks, was driven out by the ecclesiastical authorities. Tumas may still be in Vilnius for a while.

Are you now publishing your research on the kinship of the Aists and Slavs, or something else? The question of kinship is of great interest to me, too.

Little by little, the whole body of Aist studies is beginning to collapse! Now I have clearly seen that your account of Latvian intonations and of the development of ė (ie) from ei is well founded. Now, keeping Lithuanian forms in mind, I see that ie at the beginning, middle, and end of a word comes from Aist. ei under accent.

Lėpa, siena, siela, pieva: šeiva, šeima, neiva ~ dienā newly formed from dienā… Šiaul-iētis, from which newly formed vōk-ietis, all, others from visiē < Aist. *visei, from which newly formed gēdingi.

apiē, priē with ie from apiē-[linkę], priē-[taras]: nei ~ nei, bei, vei ‘look!’, la, tei, šeī from proclisis.

Latvian tris ‘three’, nē, ne k ā = Samogitian ne-kō, tei šeī have a circumflex accent, derived from the older falling accent (*tris, etc.). It seems to me that Latvian

56

In the proto-language, a sound law may at some point have operated whereby in the proclitic position a falling accent became a circumflex. How does it seem to you?

In Prussian. *tris māsas (*tris proclitic) changed into trīs māsas, just as Proto-Lithuanian *gālvā (=cf. the Žemaitian dialect of Telšiai gālvā) became galva.

I intend to finish university this spring.

I wish you success in your work!

Your Kaz. Būga.

N.B. I am sending separately an article about the dative singular and plural, from which you can make use of whatever you wish. Afterwards, I would ask you to return it to me. KB.

[2]

St Petersburg, 10 (23) March (III), 1911.

Dear Friend!

It would be a very good thing if you would undertake to write an article about the ancient “Curonians” and have it published in the organ of the Lithuanian Scientific Society, Lietuvių Tauta (Литовское Научное Общество. Вильно, Лидский переулок, д. № 7. Председатель Иван Юрьевич Басанович). So that a greater number of scholars could benefit from it, it would be necessary to write in German. Beginning with volume II, German-language articles will also be included in Lietuvių Tauta. The first German article will be Adalbert Srba’s (the Moravian Srba) “Monographie über die verba momentiva als Beitrag zur litauischen Syntax”. In writing about the Curonians, you would make use of nomina propria and this and that from Courland dialectology.

Rucava (BW. 3840, 1; Bielenst. Grenzen 231), as I learned these days from the Žemaitians of Palanga [acc. Pālanga], is called in Žemaitian not *Rūkiava, but Rūkiava (short u). BW. 12461, 3 var. Rūcava apparently has a newer lengthening; cf. Bezzenberger, Ueber die Spr. d. preuss. Letten: āsdarīt ~ asgālda 117, sūwe sūwele 118, ūs- ~ usgāldu 120. That Rūk-iava = Rucava has an etymologically short u may be indicated by the following nomina propria: Rukai – a village in the Mosėdis parish, Telšiai district; Ruk-š-iai – a village in the Pašilė parish, Raseiniai district; Ruk-ē-l-iai – a village in the Rokiškis parish, Ežerėnai district; Ruk-uiža – a surname from Liaudė (the name given to the area around the Liaudė River in Panevėžys district).

Besides Cape Ventis, in Prussia there is also a place called Wentainen (Bezzenb. Ueber die spr. d. preuss-Letten 139¹) = * Ventāiniai; cf. Jut-āiniai, Gebāiniai, villages in the Ramygala parish, Panevėžys district; Ōk-ainiai, a village in the Truskava parish, Panevėžys district; Vānt-ainiai, a village in the Radviliškis parish, Šiauliai district.

The Curonian river name Otange Bielenst. Grenzen 226 =? *Uot-angā; cf. Pal-angā on the bank of the Rāžė River (=dial. Rõužė).

Now I no longer have any doubt that in the Curonian dialect an etc. remained as it had been, and did not change into uo, as in the Latvian (=Semigallian and Latgalian) dialect. Apparently, very few remnants of the ancient Curonian dialect have survived in Kurzeme: over time, through the colonization by Semigallians and Latgalians, the Curonians adopted the “Latvian” language. We may count among the remnants of the Curonian dialect:

57

1) dzintars BW. 27421 Ēduolē; 13282 Vecpilī, Meženiekuos, Ēduolē, Irjen, Nicā, Ugālē, Dundangā, Liel-Ezerē, Kandavā are the closest neighbors of the Curonians, the Semigallians ~ Pienavā (13282 variant), Džūkstē (13282,5) = dzinterṣ BW. 27421 variant; 13246,1 Ciravā: zītars 13282,11 Liel-Iecavā (Bauska district), P. Schmidt’s = ziters 13282,3 Erlaa (Cēsis district). But why is dzintars also said in Altenē (Jaunjelgava district), BW. 13282,3 variant? Did the word come here from the Curonians together with the merchandise (товар)?

2) menčis BW. 30810 Nicā, menču rinda 18450 Ciravā: Lithuanian menkė (the dog’s tie-was wede su sawęm… menkiu meszkierioti Palāngos Juze 5₁₀) ~ Slavic mьnb: men-cis Kursiten (Ulmann) ‘a person who demands nothing’ ~ merīkas.

3) danga “corner” in northwestern Courland, Biel. LSpr. I 144, BW. 16787: 16876; 21130 Alsvāngā; 21631 Kasdangā:?garā danga ‘southwest wind’ in Nicā (Ulmann): dandzis ‘wheel rim’ (where is it used?).

4) tencināt ‘to thank’ BW. 11575 in Sarkanmuiža, Ventspils district: ténkinuos ‘I am content’ Kv. ~ iš-téko “it sufficed.”

5) devanta vasarinā BW. 8973; 12586, 8 in Sasmaka, Talsi district: where did the a come from here?

6) lanka BW. 655 in Kabile, Talsi district: luocks “laņkas.”

7) jentere Angermünde (Ulmann): ieter, ieterit! voc. BW. 23659 in Dzer̄ve, Suseja, Dignāja, Jaunjelgava district.

8) blendu, blenst in Nicā, Biel. LSpr. I 144, blenzdamies ‘while whining’ BW. 22249 in Rava, Aizpute district: blandaũ Jušk. dictionary I 215.

9) blank-stities BW. 21205 in Cirava: blānk-sterēti J. or blānk-ti ~ nu-blanko ‘turned pale’ in Marijampolė.

10) sklanda BW. 19788 in Rava, Aizpute district = pa-sklānda cf. Miež. 173 ~ pa-sklānda Ds., Užšpaliai, Jušk. dictionary I 165 b. under the entry atšlajē ~ paslāndis, -ies Joniškis “a stretch of winter road, neatly swept smooth with sled runners, slippery” (= pa-siautā Tverečius): sklendžiu.

11) the mother of the two at-linguoja BW. 23547 in Alsvānga = at-lingúoja Ds.: liguot.

12) kanc-inā-t ‘to investigate’ in northwestern Courland, Biel. LSpr. I 145 = lie. kan-k-in-ti ~ ken-t-ēti.

13) Curonian lanktes Ulmann: laņktis Kv., Kelmė.

14) denktṣ “strong” O. Bartau (Ulm.): dencis “a small, sturdy boy” (where from?).

15) Curonian (?) duncis = dūcis P. Schmidt’s ‘such a knife’.

16) guņģis ‘a bend’ in Courland (Ulm.): Lithuanian guņga ‘hump, lump’ ~ gugā, g. sg. gūgos “the same”: gauņ-aras.

17) vingr-ums in Aizpute (Ulm.): Lithuanian vingris ‘a bend’ in Joniškis.

18) vink-alāt ‘to wander about without work’ in Ēduole (Ulm.), from vi-n-k: vỹk-ti ~ vaik-ýti.

19) rantēt ‘to fell a tree’ in Dundaga (Bezz. Dial. 174): rantýti ~ ręsti ~ rāstas = Latvian ruoti in Sauka.

20) blezdelinga Bezz. Ueber die Spr. preuss. Lett. 144, if it is not simply a remodeled word based on Lithuanian blezdingā and Latvian bezdelīga.

21) Dincis proper name. Bezz. l. c. 146:? Whether vanga, venga (Ulm.) are known in Kurzeme, I cannot know.

58

Meite (P. Š[midt’as]), riñda P. Š., sence Oppekaln, skrañda P. Š., ceñšūos P. Š., balañda P. Š., bañga P. Š., treñkt are known very far from the land of Kurzeme, which makes it difficult to regard these words as “Curonianisms.” Some of these words may have n instead of m. Meite, in its accent, does not at all agree with Lithuanian meñtė Kved. or mentė, acc. meñte in Skāpiškis, Seinai; cf. a̯lu m ė s t i ‘meñtēt’ in Joniškis, Vadōkliai. Latvian mēnte (: mieturs = mentūris in Dusetos) is very probably borrowed from Finnish or from a related Lithuanian language; cf. Finnish mäntä ‘the butter churn staff.’ – Treñkt ‘to drive’ in bėgu is not to be distinguished from triek-t = treñkti? Latvian treñkt could derive from Baltic trem-k-: Lithuanian trempti ‘I drop’ ~ iš-tremti ‘to expel, exile’ J. žod. I 653: trimù ‘I tremble’ Ness. Wb. 115.

Riñda = Lithuanian rinda (Daukantas) from *ri-n-dā ~ Irish rind ‘constellatio’ Walde 436: Latvian riedu, ridu, rist ‘to arrange’ (cf. lieku, likt) ~ Serbian rěd, réda ‘row’ = Lithuanian *riñdas: rinda, -õs, riñdą (Dusetos, Skāpiškis) ‘annual rings of a tree or horn, by which the years of a tree or animal are counted’ [tree-stump riñdos = Samogitian rings = слои]: root rei-d(h)-: rei-v- > Lithuanian rievė, rievę Aist. stud. I 8738 [: Jušk. ž. I 235 ~ rivē = Samogitian rivė < rievė] Old English rāw from roi-v-: Middle High German rihe ‘row’ from i.-eur. *rei-kā ~ Lithuanian rikià ‘order, row’ Daukša: rei-g(h) > Samogitian viskas eit pa-reižiu ‘everything goes in order, as it should’ Kved. = Latvian pa-reizi: rei-s- > Middle High German reise ‘procession, journey’ from roi-sā: rei-dh- > Old High German rītan ‘to travel, ride’ and Irish riadaim ‘I travel’ ~ Lithuanian riedù, riedėti ‘to roll’ in Jūžintai = rietù rietėti ‘ditto’ in Dusetos < rei-t-. Latvian riñda can only be explained as a loan from Lithuanian (or perhaps Curonian?).

Latvian bañga (cf. Latvian buogs ‘crowd, multitude’: Lithuanian pra-bangà ‘excess’ Rupestis prabangì ape pianu. Лхр. 618, per bingusias rupestis 1. c. 8324 ~ liga į-biñgo ‘the illness intensified’ Kved.) is also borrowed from Lithuanian.

Latvian skrañda and balañda, too, are apparently loanwords. Skār-ti > sk(a)r-an-dā, bāl-tas > bal-andā, gen. balañdos and balandõs, acc. sg. balañdą and bālandą (both forms in Dusetos) [balandė, acc. bālandę in Linkmenės]; cf. Slav. želǫdb, Latin gl-and- ~ βάλ-αν-ος, Prussian alisk-ande ~ alsk-n-is in Širv., Latvian galūoda from *gal-andā, Polish łabędź from *albh-an-d-i-, Latvian til-andi ‘teals’ from a Curonian dialect, val-andā. On account of its intonation, Latvian balañda could have been borrowed from the eastern Lithuanian form balanda [with evenly rising stress on an]; cf. Latvian riñda < rindà, bañga < bangà.

I cannot explain the forms sence, ceñšūos with Indo-European -n- as Latvian. These words could have entered Vidzeme from a Curonian dialect through the Ventspils people, who lived in the vicinity of Cēsis. Trikāte speñdz ‘spiedž’ would also be a Ventspils word.

As you write about the Curonians, please review once again the Latvian dialect forms with an, etc. I repeat once more that an article about the ancient Curonians is much awaited.

Latvian piñ-ka ‘a tuft’: pīt ‘to weave’; cincā ‘kinka’ ~ cink-slis ~ cim-slis in Ulm. (if this is not an error; it would show that kinka comes from *kim-kā), kankars ~ ka-kars, cinkurs in Lubāne ~ ciekūrzis, en or eį?, dzindz-ināt < ging-, (d)zint-ele, cenceres ‘legs’ in Kurzeme, blandīties ‘to roam about’.

That Latvian kert is remodeled from *cert on the pattern of – kart is beyond doubt. The form *ceĭt, apparently, is known to eastern Latvians; cf. cēris ‘catch hold!’ BW. 31393,1.

Prussian 2 pl. es-tei 71₂₈ billi-tei 79₃₆ druwētei 67₂ laikutei 55₂₉.

59

[3]

19 17 III 11 m.

Dear Colleague!

Dignaja, as shown by Latvian eastern Dybnõja (in BW; I could not find the place this time), should be read as Dignaja < Proto-Latvian *Dubnãja [ub > ib by dissimilation]. That Lithuanian bn turns into gn is shown not only by dùgnas = Latvian dubens < *dubņs < *dubn(a)s < *dubna-s, but also by other words; cf. lùbn i os ‘ruins’ Šalakas > lùgn ios Ušpāliai: lùbos ~ Slavic lubo.

That river names, for example ones such as Lithuanian Daũgava Latvian Daũgava, Lithuanian Šešuvìs in Gaurė parish ~ Latvian Sesava BW. 381, 11; 19283, 1 var. ~ Latvian Sese, river Bielenstein Grenzen 97 ~ Ses-ile, Privatgut 1, c. 232, Pienaujā, its Pienaujā Kvėd., Riet. < *Pein-avī, gen. sg. -avijās [cf. Greek μυῖα ~ musia, gen. sg. musios Dus.: Vilkaujì, -jōs Vilkaujā, river > Vilkav-iškis, town, Suvàlkai province] = Latvian Pien-ava and others, -avā can mean ‘river’; I entirely agree, but it should be noted that another explanation is also possible. Daũg-uva = Daũg-ava could be ‘adjectiva’ with the suffix -o-vo-, -o-vā- (cf. Brugm. Gr. II², 1, p. 204: ai. kẽca-vā-s etc.). Pien-ava could be compared with Slavic dǫbrova (Vondrák, Vgl. Gr. I 410), Lithuanian welinuwá DPost. ed. Volteri 97₂₈ (: węlinuwós 173₃₃, wēlinuwą 173₃₀): šun-avā Dieve­niškis, Ašmena district (here it is not ‘life’ but šunavà), brol-avà, gen. sg. -āvos Dus. (and brolãvoj prisėina sus’pūkyti ‘even between brothers...’, Russian жидова: bendr-ovės (gen. sg.) meadows Dus. ‘common meadows = community meadows’ ~ Viēst-ovė, river in Kulių, Plungė parish, Telšiai district, Kirkšn-ovė, river in Raseiniškiai, Gin-ėvė, river in Kauniškiai.

Rivers: Bārtuva = Latvian Bārtava, Blindži uva in Salantai parish, Telšiai district, Gerd-auja in Gargždai parish, Telšiai district, Várduva in Alsėdžiai parish, Telšiai district, Mituva in Jurbarkas parish, Raseiniai district, Sandruva in Šidlava parish, Raseiniai district (if not from San+druva), Vid-auja in Eržvilkas parish, Raseiniai district, Gard-uva in Baisogala parish, Šiauliai district, Ring-uva in Kuršėnai parish, Šiauliai district, Kūrtuva in Kūrtuvėnai parish, Šiauliai district, Maĩžuva in Linkuva parish, Panevėžys district, Liñkuva in Linkuva parish, Lātava in Troškūnai parish, Ukmergė district, Vāduva in Skāpiškis parish, Ukmergė district, Mituva in Skāpiškis parish.

In terms of its suffix, Latvian Dignāja < Dubnāja can be compared with Lithuanian Lukõjus, a lake in Linkmenys parish (: Luk-nā, a river in Kulių parish, Telšiai district).

“Ni(e)granden” apparently sounds in Latvian like Nigranda|e, cf. Nigrandnieki BW. 12866, 3 var. < Nigr-ande: Iv-ande. Alš-vanga is now clear to me too: Alš- from alš(ń)- < alšņa-, or from alša- = Baltic als-ja-, whence the “Curonian” gen. pl. alkšu (rikšte) BW. 9778, 1 var. (Dundanga), if not from *alks[ń]io.

Curonian Tā ringāja Blintenieces in Gudenieki district BW. 15642 in Alšvanga: Latvian Rīga = Lithuanian rìnga (cf. svilpa < svilpti, rėka < rėkti) ~ Ringys, a river in Pašilė parish, Raseiniai district, and Užpaliai, Ukmergė district.

Curonian ar spangnem kumelem BW. 14510, 2 in Dundanga: spāngas ‘rags’ Suv. Naumiestis.

Curonian siva, centra meitiņa BW. 12314 var. ‘dzedra’ in the Sasmakas area, Talsi district:?

60

Įlendams 12785 BW. 607 skrundā kurš. meitas lenca BW. 11985 Alšvangā: lankvīti (= la. luocit) ‘навецать’. {vilk(u) māt(e) jēr(u)s lenc „kerigūsta“ BW. 23566,4 Rinde (Angermünde) kurš. Dzelde (Emboutes par.) > Dzeld-enieki BW. 12684, 1 var. rodo, kad lie. gēlda „корыло“ yra savas žodis. Tuk-ums, plg. la. Jaudz-umi, māju vārds BW. 19079; 18314, kurš. sumi…ieraudžija… nuo-skrendušu kumelīņu ‘nustipusi, nudvėsusi’ BW. 19071 Vec-Sieksātuos Āzputes ap.:? skrōnda. kurš. rindāt ‘maut = murōti ~ maušti’ BW. 16518 Ventspilī, Vērgale, Rāvā:? Dulenka (gen. sg.?) strimalu kupči BW. 20724 Cīravā Āzputes aps. ~ māju vārds (?): la. suf. (e|in)-iek(a)-s < -enka-. Tinger-nieki BW. 20949 var. Dundangā < *Ting-ure: tingėti. Dundz-inieki BW. 20949 var.; 20957,3 Dundangā < *Dunga:? vīhante BW 21871,6 var. Ventspilī = vibuot-ne 21871,2:? planda kārta? (reikšmė nelabai aiški!) BW. 23583 Alšvangā, Ventspilī:? rank-ainis kakliņš BW. 25794,2 Kabilē Talsu ap.:? vok. Runzel < rṇ-d-, la. ruōbs < ran-ba-. Ri-n-dā ~ ri-k-idā ‘eilā’ tolesniai yra giminės su lot. rī-t-e^* ‘in der Reihe’ loc. sg., rī-tu-s, ahd. rī-m ‘Reihe, Reihenfolge, Zahl’ Brugm. Gr. II², 2, p. 710: visi šie žodžiai paeina iš šaknies rei- ‘tekėti’ (: saule teka, vanduo teka, merga už vyro teka, tākas, tekmē „лужа“) ~ slav. rě-ka, lot. rī-vus. Iš Rozwadowski’o la. nėsanti lytis *ē-dzēris (= ie-dzēris) jau pateko į Brugmann’o Gr. II², 2, S.817, iš kur, galima spėti, pasplis ir tolesniai per žmones, nemokančius latviškai! Locat. sg. vilkė galėtų būti iš *vilkēn: vėd. Varuṇān-ī f. ‘des Vāruna Gattin’ iš loc.; cf. Agnāy-ī ‘des Agni Gattin’ Brugm., KGr. 384, Manāv-ī ‘Gattin des Manu’ eigl. ‘die beim M.’ Brugm. KGr. 330, 384, Gr. II², 2, p. 514 ~ ai. sūrē duhitā Gr. II², 2, p. 514: Naujokie-nė ‘Naujōko pati’ < loc. *Naujokie (cf. svāi-nē ‘свояченница’ iš loc. sg. *svoi ‘bei sich’: ahd. swei-n ‘Knecht, Sohn’, av. XVᵃe-paiťis ‘selbst’): žem. Paknēnē ‘Pakniės pati’ < loc. sg. Paknē(i) [nom. sg. Paknis]. *Vilken (cf. Varuṇān[-ī]) būtų iš vilke [основа = kamienas] + en; cf. aind. kṣām-an ‘auf der Erde’ Brugm. Gr. II², 2, p. 727??

I wish you much success. Kaz. Būga

[4]

St Petersburg 19²³⁄^III 11.

Garbus Drauge!

Thank you for the three new installments of your studies! I find in them much that is new and worth knowing, too. Your conjecture that not every ēu turned into jau seems quite sound to me. To resolve the question finally, one would need a complete list of words with jau and au. It is true that forms with jau are far fewer than those with au. To pjaułai (: accus. pjāulus Vilkaviškis = pjaułūs Kvėd.), kiaułai (: acc. kiāutus ~ nom. sg. kiāutas Dusetos, Kvėd., Suv.

61

Naumiestis) galima būtų dar pridėti: pliaūtas, plur. pliaūtai Tverečius ~ pliaūtas, pliaūtai Dus. „должен = avilio lenta, durys“, žem. šiau-t uvas „ткацкий челнок“, liáunas, -à „гибкий“, напр. liauni pirštai Kved., Šiauliai, Joniškis = la. launs, žiaunos, -ų Dus. „скулы“, liaukà, -õs liaũką „волна, didelė banga“ Joniškis, šiaurės-vakar. žem. žiaudùs „суровый“, jāura „juodžemys su moliu = pilkoji žemė (= šal-

soil)” Dusetos, Tverečius, Linkmenys.

Сл.-балт. этюды 86: Samogitian dėves jau K. Jaunius explained as deriving from *de-d-vēns = cf. Sanskrit dadvān; see Aist. stud. I 64*.

стр. 90 Laidaras [: išz laj daro Dauk. Būdas 15₂₇, laidars l. c. 22₃₁, cf. Jušk. žod. I 249a, Dauk. Lietuvos istorija I 189, II 169: laidars Jušk. žod. I 310 b] is attested only with a circumflex: laidaras in Joniškis (Aist. stud. I 75). Laidaras is known only in the Latvian borderlands. The intonation suggests that Lithuanians borrowed this word.

стр. 90. Prussian pelanno ‘hearth’ I explain from Proto-Prussian pelenā: Latvian out of le < le, cf. eastern Lithuanian pelenai ‘ashes’.

Prussian assaran from Proto-Prussian ezeran: eastern Lithuanian (in the Ashmyany district) āžaras alongside žāmė ‘earth’ (= Prussian same V. 24)

Prussian glosano V. 778 ‘gluodenà’ cannot be emended to glodano, since a reading error s for d is hardly possible. I read glosano: from Proto-Prussian glās-enā; compare eastern Aukštaitian šānas < šēnas alongside dedù vedù nešù (the town of Dusetos, Šalakas, etc.): Prussian assanis V. 14 = ‘autumn’.

Prussian pelanno from pelenō, cf. clattoy V. 292 < Middle High German klette, ladis V. 56 = lād as in the town of Dusetos < lėdas in Pažiegė, Dusetos; lagno V. 125 ‘liver’ = Baltic *leknā Pedersen Kelt. Gr. I 129, 177, lattako V. 543 ‘ice-hole’ = Baltic *leda-takā, melato V. 743 = meleta ‘зофна’ in Širvydas’s Dictionarium 457, pelanne V. 37 = pelenai Dusetos < pelenai > Samogitian pelinai Kved., pelanno V. 223 = Samogitian pēlins ‘hearth’ Kved. pēlenas, plasmeno V. 148: plāsmas ‘kindling’ Dieveniškės (Ashmyany district) < plesnas = Latvian pleksna ‘the foot of a stocking’.

Pr. au-klextes V. 280 klexto 333, clenan 194 (: klē-tis), pellekis 202 ‘Giebel = acies tecti’ [: vežys ištięsęs pelēką plaukia Jūžintai: pēlakas Dus. < pēlekas, pele-maygis 713 (cf. petē-kautai ‘pelėms kauti prietaisa=sląstai’ Dus.), pleske 253 turi e iš ē.

Calene V. 231, kele-ranko 303, sylecke 575, scalenix 703 have e from ī: kalinē, keli-rankā [cf. nami-šōstas ‘estate’ Linkmenys < nama-, blauzdi-kaulis Vadōkliai < blauzdā-, šōnikaulis Vadōkliai šāna-, Prussian Crixti Lāiskas < kriksti-lāiskas ~ kriks-ta- ‘baptism’], sili-kē (: Slavic slb-db), skal-ini-k(a)-s.

Wilenikis V. 439 should be emended to Eidenikis = eid-inika-.

From what has been said above, it is clear that in the Elbing dialect le changed to la (= lā). In the Elbing dialect, not only l but also r, s, and z unexpectedly hardened before e.

rē > rē > r'a we have: kraclan V. 118 = eastern Aukštaitian *krākļai < krēklai ‘breasts’, raples V. 520 = rāplās Šalakas < rēplēs Jūžintai, ratinsis V. 368, 540 = Proto-Lithuanian rētežb. Drawine V. 393 ‘wooden beehive’ could equally have arisen from Proto-Prussian drey-inē [:drēvē ‘tree hollow’ Kvėdarna, Joniškis, Salantai] as from drav-inē: Lith. draves nom. plur. BW. 18360, 2 var. ~ dravās kuoks 15729.

62

se > sa: assanis V. 14 Russian осень, possi-ssawaite 20= Lith. pùs-sevaitė: Semeno V. 752 (= sėmeną Debeikiai) and seveynis V. 229 ‘pigsty’ have a long ē. Proto-Prussian sēv-eina-s ‘suinus’ [cf. Lith. lap-iēnē subst. from adject. lap-eina-]. Sēv-: sū- ‘ʒ’ = bhrēv- ‘Brane’; bhrū-.

Nose-proly V. 86 has e from i: nozy 85 ‘nose’ ~ eina nósį-rakas Dus. ‘scraping the ground with its nose, stumbling’. Wissene V. 622 ‘rooster’ from *vis-inē: Skt. višá- ‘poison’.

The Elbing dialect forms ladis ‘ice’, pelanno ‘ash’, raples ‘tongs’, glosano ‘glossy snake’, assaran ‘lake’, same ‘earth’ differ in no way from the forms in the eastern Lithuanian highland dialects of the Ašmena and Trakai districts: ladas, pelanai, râples (ragēt < regėti, râtas < rėtas), sánas (< sēnas. In my dialect, sėnas, but sanai, sanātuvė, sanýbė), ažaras, žãmė. The Prussian dialect of Elbing closely resembles Lithuanian dialects in the far east of Lithuania not only in that le, re, se, ze change into la, ra, sa, za, but also in that word-initial e- changes into a-.

e- > a-: 1. addle V. 596 = ãgla Joniškis ãglā Linkmenės < ēglā Dus. town < ēglė Jūžintai = ēglė, 2. asy V. 241: ažìa, gen. sg. ažìas Linkmenės ažìa, ažìos Vilkija < ežia, ežios Dus.; 3. assaran V. 60 = ažaras Dieveniškis, Merkinė, až’aras Linkmenės < ēž’eras Dus.; 4. asseges V. 572 from *azigis: ažgỹs Linkmenės from ežgỹs. In Salakas they say both ažgỹs and ežgỹs ‘asp’.

Abse V. 606 ‘fir, -’ Dus.’, aliskande 602 ‘alder’ Dus., aketes 255 ‘harrows’, arelis 710 ‘eagle owl’, assanis V. 15, aswinan 695 ‘mare’s milk’, alkunis 110 may also have Aistian a-.

Ar-warbs V. 301 ‘crossbar’ could have ar- = Lith. ar ‘with’ (: Lith. sa-varas: durių varai ‘door leaves’ Kved.). -war-b(a)-s ‘-varas’: vāras ‘pole for making a fence’ Dus. durių vāras ‘door = leaf’ Kv. = pa-skāl-ba ‘the rumor’ Kv.: pā-ska-las ‘same’ Dus. -warbs is not = virbas (erroneously Aist. stud. I 139⁹).

Instead of alne V. 647 ‘animal’, together with Jaunius I read alue (cf. Nesselmann Thes. 5. + Preartue V. 249) = alvē. Alvē is a word of Finnish origin: Finnish alävä ‘living, alive; animal’ (~ elo ‘life; livestock’). We have a second word of Finnish origin in Prussian in the form kal-ēt ‘a very thin drink (beer)’ Nesselm. Thes. 62 (a borrowing from Old Prussian into the dialect): Finnish kalja ‘weak or thin beer’. Alue derives from *elvē.

Aloade V. 541 ‘hasp’ is unclear as to its origin; if it is related to Latvian el-vede ‘el-vytė’, it should be emended to alvade (for v cf. vdro V 667 ‘otters’); cf. Lith. al-varas ‘crossbar’.

Emelno V. 646 ‘dodder’ and esketres V. 567 ‘thorn’ have e- instead of a- for an unknown reason. Should it be read *ēmelnā? The quality of the vowel e in the word form emalas used by Širvydas is unknown. But cf. Latvian āmuls ‘mistletoe’. Esketres = eskētras must surely have e generalized from an unattested form *erškėtis = ērškėtis Kved. ‘thorn’?

Ennoys V. 158 ‘fever’ should perhaps be emended to eynoys: from ein-āis-a-s ~ Russian иней; for the suffix cf. caymoys V. 105 < *camoys = kam-āis[a]-s: Latvian kamiesis Lidén Arm. Stud. 30 < Aistian *kam-eisis.

Estur-eyto V. 776 ‘lizard’: ēstūr- from *jēstūr- [cf. eristian V. 681 from jēra- = Latvian jērs = Lithuanian jėras ‘lamb’ Lazdūnai Dieveniškis] = Polish jaszczur (cf. борщ ≥ Lithuanian barštis).

63

Estūr-eitą sufikso žvilgsniu galime palyginti su: pra-šal-iėtis „Fremdling“, gimin-iėtis ‘giminaitis’.

Penpalo V. 770 ‘blowfly’ I emend to peypalo [np would have turned into mp] and read it as the same word as Lithuanian piepala Baisogala (cf. Aist. stud. I 82²⁴). The Aist. form pei-palā derives by reduplication from the root pel-: pelē ‘мышь = *серая’ Greek πελιδνός πολιός. For the degree of reduplication, cf.: vie-versys, vie-versio Debelkiai, Panevėžys ‘lark’; žie-žulā, -ōs žiežulą Dus. Linkmenys, “red-finned [fish] Tverečius” (such a fish); žie-žirba ‘spark’; Sie-sartis, iēs Siesartį ‘the little river Cesarka’ in Kaunas, Vilnius, and Suvalkai provinces (there are many of them); Latvian Cie-cere, a river at Kuršė (: Kerēs-upis, a stream in the Užpaliai area), Vei-viržas, a river in Raseiniai County. Latvian pal-pala may be compared with gai-galas, vai-vōras ‘cranberry’ (such a berry), žai-žar-ōkas Jušk. dictionary I 649 b. Proto-Slavic per-pelo-has an analogous reduplication alongside Prussian werwirsis V. 733 ‘squirrel’: ve-versys Vadokliai: vo-versys, vóversio Vabalninkas ~ voveršis Dus.: vie-versis Linkmenės [here pronounced *véveršis, as also dēveris alongside acc. dieveri], var-vōr-ykstē ‘rainbow’ Butrimonys ~ vai-vōr-ykštis Kv. (for the suffix, cf. šeimyn-ykštis, band-ykštis).

стр. 92 švendrai = švendrai, acc. švėndrus Vilkaviškis: švėndrė Dus. (en > in).

p. 97 žāsis ‘goose’ I have heard with my own ears some 10 times in Zasėčiai (the village of Zasets, Dyatlovsky < 16th-century Zdetela, Slonim district): žāsis, iēs nom. pl. žāses, g. pl. žasū. Zasėčiai is also a Prussian name; cf. Lithuanian Жосама, a river near Veliuona (for the suffix, cf. Vės-etā, a river flowing past Upytė in Panevėžys County, Vēr-etos, a river in the Ylakiai area, 13th–17th-century Dus-etā, a river now called Šventoji near Dusetos).

Pr. Weware = vēveris vēveri Taurāgnai (Aist. stud. I 62¹⁵), Linkmenės ir vietomis Dusėtiškiai (mānā dēveris < *dieveris kaip vēveris).

стр. 99 *drēvis with long ē is unknown to me. Drēvē in Kv. Joniškis and Salantai: drovē drōve ‘hollow’ in Skāpiškis.

стр. 101 la. gliēmēzis may be compared, on account of its ending, with Lithuanian glēmēžius in Dus. ‘slime’ (silkės aptrūktos glēmēžium).

Стр. 104 Pr. naw-an(a)-s I Kat.: neuw-en-en II kat. = la. ryt. slyd-ōna- BW. 13597, 1: la. žem. slid-ena BW. 13597.

стр. 109 la. šlau-b-ums „отлогость“: + pa-šliau-nuš „ukosný“ Jušk. žod. I 557 a ~ la. liki ragi, šlauni sāni BW. 28923 slauni sāni l. c. var., slauniem sāniem i. pl. BW. 28956 < šlau-na- [> Šlaunis, upė Rietavo par.]?: kle-v(e) ~ k’le-j(e) > šliei šlaitas: kel(e) > ahd. hald ‘geneigt’.

p. 111 Latvian to struggle. The same is meant by Lithuanian to exhale in Tverečius.

p. 121 Latvian klaips. Sirvydas’s kliepas ‘loaf of bread’, which cannot be borrowed from хлеб, shows that Latvian klaips too could be a native word. Finnish leipä is explicable only from Lithuanian kliepas, but not from Germanic hlaiba-

P. 125 Plikšas not from *plik-sa-s, but from plik-šas, cf. varg-šas, gōb-šas, gūd-šas > gūčas, stit-šas > stičas, gint-šas > ginčas, trīd-šas > tričas, drib-šas etc.

Стр. 16 dumjū < dm-. Dumjū gali turėti ir i.-eur. u; plg. ϑύελλα skr. dhū-ka-s ‘Wind’.

стр. 18 Alongside eel (cf. Prussian angurjis) we also have ingurys in Dus. Kupiškis (Aist. stud. I 109), Vyžuonos (in may also be from en).

64

p. 19 grumėti ‘to thunder’ Kv. may have u also from Indo-European u; cf. grūmoti ‘to threaten’ Kv.: grū-džiu, grąusti ‘to thunder’: nu-gramėti ‘to fall down somewhere with a crash’ ~ βρομ[ῆ]: garmė-ti ‘to go with noise and a crash; to plunge into the abyss’ = *βρομῆ: grum-džiu ‘I threaten’.

p. 32 vėveršys: vė- is entirely unattested in living speech (see above). malatà Dus., Linkmenės, Jušk. žod. I 426ᵇ ‘green woodpecker’: meletà Kved., Kurš., Širvydas = monomo: tenemo = gàrbana ‘curl’: *gerbenė > gerbenėtas Dus. ‘curly’.

p. 91 vėšelas is a non-Lithuanian word; it is borrowed from Russian вёшало ‘frameworks, fences of various kinds for drying something.’ Borrowed from such a Belarusian dialect in which š changes to s: vėšalo > vėsalo > Lithuanian vėsėlas. For š, ž > s, z, cf. borrowed words: pažaras ‘fire’ pažytkas ‘benefit’ < пожитокъ.

I wish you all the best, Kaz. Būga

[5]

Mósėdis (Old Lithuanian May-sedis, see Bielenstein, Grenzen…)

December 26, 1911

Dear Friend!

Thank you for your little letter! I thought you would have left before Christmas. I assumed you had been in Vilnius: “Viltis” had announced that you would be giving a talk on the Baltic languages. I received your letter with Mülenbach’s address. I am at fault for having kept silent until now.

Jablonskis’s address: Гор. Брест-Литовск, мужская гимназия Преподавателю И. О. Ябл.

I too now doubt that ps changes to pš (cf. vapsà ~ vapsvà, gen. sg. vãpsvos Dus.). It seemed to me that Dus. ėpušis, gen. sg. ėpušies (ėpušė is also said. Cf. šėrdìs, -iẽs, šėrdį and šėrdė Dus.) derives from *ėpšis, gen. sg. *ėpšeis. *Ėpšis acquired u from pušis, iẽs. Ėpušis, iẽs could also derive from *ėpušis, iẽs, with š from pušis. Gen. pl. ėpuš-ų (~ puš-ų) and ėpušių. ėpuš- from ėpuš- (following puš-) relates to eps- (cf. Latvian aps-e) as els- (elks-nis < *els-nis, Russian лешина < els-) relates to elus- (cf. Germanic aluz/s- Russian ёлха < *eloxa).

The Anykščiai people say kl’ėvas (with a broad e), just as they do jėvas, jėknos, kėvalas. All eastern Highlanders distinguish l’a from le in suffixes: kėlias, but acc. pl. sáules or sáulas. K’elīavo = Anykščiai, Kupiškis keliōᵃvoᵃ. mėzliava ~ gãniava or (elsewhere, Dus.) mėz-l’eva ~ gãneva, alongside devýni (hard d. Dus.). The people of Kupiškis, who pronounce apjōᵃkoᵃ ‘I went blind,’ do not say *kl’ōᵃvas. The Kupiškis form kl’āvas or kl’ėvas (I cannot say today which it actually is) can only be understood from kl’ėvas: ā began to change to ō when l’a had already changed to l’e. Instead of apjōᵃkoᵃ, *apjėko, with oᵃ from *ōᵃkoᵃ. Apjāka (Dus.) with a from āka or from the early-Dusetiškiai present apjankù (> *apjunku): easterners knew how to distinguish jan from en: vėrpja-n-ti > vėrpjunt Dus. vėrkja-n-ti > verkjunt Dus., but sarbintà from serbentà ‘blackcurrant.’

Alongside apjāka, the people of Dusetos say jėvas, jėknās (some also say ėknās, even āknās): e is neither broad nor narrow. jėvas and jėknās may have Indo-European e. Alongside kl’ėvas

3. 599

65

The people of Dusetos say only kel’āvā (važ’āvā etc.), where ’ā stands in place of the expected e from the infinitive kel’āutē (t of intermediate hardness), future kel’āusū. The acute ’a does not turn into ’e: jai d. s. fem., jām d. s. msc. ‘to her’ ~ ‘to him’ in Dusetos, arkliām d. pl. ‘to horses’ in Jāi, but vāl’ei in Dusetos. Accus. pl. val’as, dvasias, nom. sg. valia, dvasia in Dusetos. Gāneva in Dusetos (n is soft).

“(There is no one who) dāra” is the nom. pl. participle from daras. Cf. “there is no one who dāras.”

The people of Linkmenys do not say Ciltas: this is said around Kaltanenai, and t. āgla is ’āglā: both a’s are of equal length (средней долготы). Prāšam is not equivalent, and neither is the Linkmenys form “mes (būlbas) kāsam ‘копаем’: prāšām (ā of intermediate length), but kāsām.”

The Linkmenys gen. sg. širdēs, as shown by Tverečius parish širdēs (alongside pìrčiēs “bathhouse”), may have the Baltic -es; cf. Daukantas’s širdēs.

I know nothing of what Bezzenberger says about Trautmann’s grammar: I have not had a chance to read this review.

Such adjectives (adjectiva) as Latvian led-uoja- ‘icy’ BW. 13268¹, 25 var., celm-uoja- ‘stump-covered’ BW. 19604 var., dūm-uoja- ‘smoky’ BW. 14594, naiz-uoja- ‘unmarried’ BW. 16278, diev-uoja- ‘deivōtas = good’ BW. 12409, 1, were not unfamiliar to Lithuanian in the past either. For the time being, I find the suffix -uoja- only in the place name Kaln-uoja-i (Prussian Lithuania) = and Žemaitian Kaln-ūjai (Raseiniai district, where they say dūna < dūona).

Kaln-uo-ta-s: *Kaln-uo-ja-s = apyn-ō-la-s ‘hop garden’ in Kvedarna and Rietavas: apyn-ō-ja-s ‘the same’ in Mósėdis and Salantai (dialectally: apynuos, gen. sg. apynioję).

If ruinis, ruišis come from Indo-European *rōinos, *(w)rōikōs, how then are we to understand vārna alongside vārnas? Until now scholars have maintained only that vārna derives from Indo-European *wōrnā (but *wōrnōs); cf. vilkė alongside vilkas: Indo-European wǝkʷyē, but wī̆kʷos.

Cf. zuikė ‘female hare’ beside zuikis. Latvian astuotais, as shown by the Telšiai Samogitian aštōnc (ōn not quite long), gen. sg. aštōnta (ōn fully long), must have come not from Proto-Latvian *astuontas, but from *astantas. Telšiai aštōnc (главное ударение на первом слоге. Mósėdis, Skuõdas, Salantai), as shown by dūmblōs “mud,” sunkōs “heavy” (the first stress is stronger), can only have arisen from *aštantas; cf. rōnkā, acc. sg. rōnka, dąngōs ‘sky’, krōnta “берег, kraštas,” pa-šōnkōs ‘acorns’.

Samogitian aštāntas together with Latvian astuotais shows that Indo-European “ō + n + Cons” must already have changed into “ō + n + Cons” in Proto-Baltic, whence “ā + n + Cons.” Latvian dialectal *astūtais and Lithuanian standard aštūntas are innovations from astuoni = ‘eight’. Aštāntas + aštuoni > *aštuontas > *aštūntas > aštūntas.

Vilkui < *wī̆kʷōy: vilkais < *wī̆kʷōys = tū ‘τῶν’ < *tōm: aštāntas < *oktōntōs. That Sanskrit vṛkais should be *wī̆kʷāis cannot be demonstrated. There is insufficient basis for interpreting λύκοις as an innovation on account of the alleged Baltic *vilkais < *wī̆kʷāis.

Aštāntas together with vārna shows that ruinas, ruišas cannot have ui from Indo-European ōi. Could Latvian gen. pl. tuo, *vilkuō > vilku perhaps have arisen only from Proto-Lithuanian *tan *vilkān? *Tan ‘τῶν’: tū < *tun < *tuon = “these men” in the dative sg. at Debeikiai and Vabalninkas (Jaūnius): vyrui < *vyrui.

Does Latvian tuo ‘τῶν’ come from *tuon, with n disappearing before uon shortened to un? Latvian dialectal astūtais would indicate that uon was shortened before an, un, etc. changed into uo, ū, etc. How does this seem to you?

I very much look forward to the article about the Curonians. This question is of great concern to me, and I myself have collected a little material on it. Valtaiki etc. may stand in place of *Valtaīti (from *Valtaiši, as also in Samogitian …).

66

jaunėkātė ‘young men’). Do the Latvians of Courland, alongside pairīt ‘the day after tomorrow’, also say muims < mums (Samogitian acc. pl. mōmìs, i. pl. mōmìs: i is close and m before i is palatalized). Do you regard words such as Latvian sklīdēt as Courlandisms (cursismus)?

One or two words with ui among the Telšiai Samogitians could be remnants of Curonian. Cf. paukštuitis in Mósėdis, Ylakiai. Kuisis ‘mosquito’ (-i- close. Kártės ‘kārtis, ies’—i open = ē) in Mósėdis, Skuõdas, Salantai = Prussian cussis. Ness. Wb. 302, in Prussian Lithuanian, has puikỹs from Klaipėda and the Curonians of Kópai (Nehrung)? = pūkỹs in Tilžė = pūkis < puõkis m. in Salantai. Prussian Lithuanian rūinis ‘seal’ < Curonian *rūinis of Kópai and Klaipėda < *ruoʹnis = ruonis. Šliuizis РфВ. LXVI 231—Telšiai Samogitian < Curonian *šluʹzis, gen. sg. *šluʹža = *šluzis: šliužas ‘skate’ Ds.

I am not sure that all these words with ui are “Curonian” remnants. Kuisis is also said around Saločiai (Panevėžys district), on the Mūša, not far from Bauškė. With paukštuitis one might compare the name of the town Meszkucie (Šiauliai district, directly south of Mintauja)—Meškūčičiai < *meškūitis ‘bear cub’. Puikỹs, rūinis, and šliuizis could be Curonian words: they are known only to the neighbors of the Curonians.

There may indeed be remnants of the Curonian language in the vicinity of Mosėdis, but I have not yet managed to understand them: I rarely get to meet Samogitians. I want to come next summer. With more free time, I will learn more. There are several distinctive dialects here. In Mosėdis they say: ōžūols ‘oak’ (the second ōu is falling—gestossen—and almost as long as the first), žõušės, gen. sg. žõušėis (the first stress, though short, is more prominent) ‘goose, of a goose’, žõušū ‘of geese’, vỹrų ‘of men’ (u is close; but dô vỹrō—u is open), kõu ‘what’, ānõu ‘that’.

In Salantai they say: ōžouls, žõsės (ō is almost long, but not as long as in the accusative žõsi: i is close), gen. sg. žõšėis (ō is of medium, if not short, length), žõšīm, vỹru (but dô vỹrō), kõn, ānõn.

The Telšiai Samogitians have two kinds of short u and i: close (узкое) and broad, or open (открытое) [I write: u and ō, i and ē]. The people of Mosėdis and Salantai say:

sūnõs nom. sg., but dô sūno, tris sūnus acc. pl.; dõgnõs ‘bottom’ (ō of medium length), dô dõgnō, tris dõgnùs; tõ ‘ты’, sõ ‘сү’, c’ nõ ‘nunc’, tūs ‘тех, туõs’, ānùs acc. pl., šiūs acc. pl.

šēs ‘съ’, but ši ‘си’; āķēs ‘eye’, but āķis nom. pl., āķis acc. pl., āķi acc. sg., āķē instr. sg., dvē āķē

āš tōriō, tō tōri (reflexive tō tōrỹs)

ìt, ìkì, lìg, nō-gì or nō-gìs ‘nunc’

The close u and i derive from Baltic ū and ī.

tūs vỹrus < Baltic *tūs vỹrūs

vẽšus sons < višūs sons

(dō) sūno is a newer formation by analogy with (dō) vẽrgō; cf. nom. pl. sūnā (ā is longer than ū), as in vergā.

Acc. sg. sūno is a newer formation replacing sūnu, as Daukantas stated. Daukantas always writes: sunu, alu, midu, medu, amžiu, but dô sunō, medōmi, etc. Accus. sg. sūno has ō instead of u, following i. sg. sō sūno, nom. and acc. du. dô sūno. Alongside accus. sg. āķi there is i. sg. sō āķē [the first stress shifted onto the preposition], nom. du. dvē āķē.

3*

67

tō, sō, nō are Baltic tū, sū, nù. ši = la. ši: Greek ἰ Slavic cu.

but šės = la. šis

ãkis n. pl. < *ãkis, akis acc. pl. < *akís, āki acc. sg. < *āki < āķi. sō akê = Dusetos. sū aki < *aken (following sun, mouse).

(dvė) ākė from Baltic (dvei) *akei in place of *akī, by analogy with *dvei peleí.

tō tōri < turī: reflexivum tō tōrýs. Alongside sūno, he also says sūnomi < *sūnū-mí.

Acc. pl. and i. pl. mòmis jōmis < mumīs jumīs: eastern high Lithuanian mumi jumi acc. and i. pl. I. sg. mōnė-mí, i. pl. trėmės ‘with three’.

Acc. sg. mōnì, tavì, savì = eastern high Lithuanian (Dusetos, etc.) manì, tavì, savì < manī etc.; cf. acc. pl. mumi ‘us’, jumi ‘you’.

Lig by prothesis from lýg.

it through assimilation from *ēti (cf. iti-n ‘good’) < *itī (cf. Kvedarna dialect tý-toks ‘such a one’), as iki from ēki < *ikī.

nō-gi cf. Dusetos argī < *ar-gī.

In the Samogitian proto-language, word-final long vowels still included not only ū, ī, but also ā, ē. Samogitian vārna, but nom. sg. neutr. pėrmō ‘formerly’, = Dusetos pirmà, mōn čė līnksmō (not *līṅksmō, if it were a u-stem word), alongside nom. sg. fem. līṅksmà. Nom. pl. kãtės from *kãtiės (cf. saulė ~ pėliė), but acc. pl. kãtės < *katēs.

Acc. sg. vārna, šãka, but i. sg. vārnō, šãkō. Cf. high Lithuanian vārnā|ū: vārnā|ū.

-ēn, -ēn uniformly yields -ē: saulė acc. and i. sg., pėlė acc. pėlė i. sg. In the accusative singular, Telšiai Samogitians have -ē rather than -e because in trisyllabic and similar words every -e turned into -ē (open -i). Cf. voc. sg. výrė, Dėivė; estė 2 pl. pres. (in Valančius) nebė-bōva (ō of medium length).

From these few strokes, Your Honour may infer how important the Telšiai-area dialects are for comparative Baltic linguistics.

In Mosėdis they say: kėistė ‘to endure, to suffer’, tėistė ‘to spread, to stretch’ and tėistė ‘to pull, to continue’.

In Salantai they say: kėistė, tėistė, but tėstė. Kėisōs, gen. sg. Kėisa ‘a bump, a knot’ Mosėdis = kėsōs, kėsa Salantai. tėivs, fem. tėivà Mosėdis = tėvs, f. tėvà Salantai = tėnvs, tėnvà Kvedarna = la. tievs.

The Samogitian dialect of Telšiai calls the Latvian land Kōršōs, g. sg. Kōršā, and its inhabitants kōršėninks, plur. kōršėninka. Kuršas ~ kuršininkas.

The Samogitians of Mósėdis and Salantai know the following Kuršō towns: Grāmse, gen. sg. Grāmzda = German Gramsden. Nygrōnc, gen. sg. Nygrōnda = “Niegranden” Vūožpūtis msc. < *Ōžputis < Proto-Samogitian *Āžputis = Cour. Āz-pūtis ‘Aizpute/is’. Buotė, manor = German Bahten. Vėrga < Proto-Samogitian Virga = Cour. Virga Pāur-ōpė, town ~ (part of Rucava?) Lėipūojė, loc. sg. Lėipūojė = Cour. Liepāja The river at Mósėdis is called Bārtōva < Bārtuva. What intonation does it have in the Couronian dialect?

The name for Riga among the Samogitians of Telšiai, Riṅgā, gen. Riṅuōs, acc. Riṅga (in the accusative the vowel is longer than in the nominative), is evidently a Couronian word. Couronian Riṅga > Samogitian Riṅgā. The Semigallian Rīga > High Latvian Rygā.

68

Kuldiga “Goldingen,” even in the 16th century (see Спрогис словарь др. жемойтской земли), was still called Kuldinga in Lithuanian.

Mockвá becomes Maksvā: sk+Cons. > ks+Cons. -isk- > -išk- in Lithuanian (Dus. Mósėdis) (В Сл.-балт. эт. 54 сл.): twjβka “it flashes, shines” Mielcke L.-d. Wb. 304, blyβkėti “sparkle, shimmer, gleam” l. c. 129, wikβwa “red small grass, fine grass” l. c. 329 ~ wikβris m. ‘cut grass’ l. c. 329 from *višk-và ~ *višk-rìs. Dus. viksvà is a new formation (~ *visk-vā: *viskā = vaps-và: vapsà). *Viskā stands in place of *viškà, with -s from the neighboring feminine form vizgā Kvėd.

After other vowels (not i), sk does not become šk; cf. vóskurnoti, éškà, préškas, pléškė, maskatuoti, láškana, lùškis etc.

-sk+Cons. must have changed into -ks+Cons. This -ks-, like Indo-European -ks-, must have changed into -kš-.

Tràškana Dus., with š from trekšti < *trēks-tei < *trēsk-tei (Polish trzaska).

Pūkštė pā-pauškas (: Polish pysk) ~ pūkšt < *piūsk-ti

Taūkšti “to chatter”: taūzyti

tėkšti ~ taškas. trūkšt ~ truška 3rd pers. pres. (τερύσχω) pliaūkšti: Polish pluskać.

Šermukšnis msc. may derive from Proto-Baltic šermusk-nis: Slavic -ъха < -us-ā. Dus. ališk nys ‘alder’ can only derive from *alis-nis: from alisk-nis it would be *alikšnis.

*Alis-nis ~ alder: Prussian alis-k-andē = Slavic čermъx-a < -us-ā: *šermus-k-nis = úos-is: Germanic as-ka-.

Rezgū ~ rēgztis m. cf. Latin res[c]-tis. Mezgū ~ megztuvė Ds.

Veizd-ėti (cf. der-ėti > der-lùs, tū-k-ti > tuk-lùs, jemù > *pa-jem-lùs > pajemb-lùs Kv.): *veizd-lùs d before l changes to g (Polish handlować > Dus. angliavóti): *veizd-lùs > *veizglùs. Through metathesis, veizglùs РФВ arose from *veizglùs.

Slavic -skn- > -sn-, whereas Baltic has > -ksn-, whence -kšn-: Бльс[к]нѣти ~ Lithuanian blikšnóti, does not [the sheep] rid its fleece, or does the donkey rid itself of its hair? What a nuisance!

Zogr. Leskien Hd.⁴193 = atā-rišā Ds.

разд-рѣшити… отъ жзѣ = iš-raišyti. Cf. Slav.-Balt. et. 69¹.

Rus. муж-ило: муж-[б]ло РФВ LXV 402 = Prussian krup-eyle: s.-Lith. deiv-ila-s ‘little god, idol’.

Church Slav. агн-ицъ: агн-ыць = Polish dialect pal-ic: pal-ec, palca = jaun-eika ‘younger brother’: jaun-ikis = jūod-eikis ‘black blood soup’ Linkmenes: juod-ikis ‘name of a coin’ = Nam-eikis surname: Nam-ikas ‘the same’.

Br. коро́ба: вр. ко́роб = ворона́: во́рон.

Skr. kljūna ‘kind of hook, clamp’: kljūn ‘something bent, crooked’ Berneker SEW. 529 = kl’ēiva ‘bow-legged’: kl’ēivas adjective = kvaiša ‘fool’: kvaišas ‘foolish’.

Russ. воло́ца ‘a person wandering about idly’ Dal’ I³ 581: волоку́ = Latvian luõža: Lithuanian lendì.

Bėgėtė (to run) cf. δρόμῳ θεῖν ‘to run’, бегом бежать is 1st sg. from *bēg-tē (the people of Dusetos say bėgti bėgu); cf. Širvydas (Dictionarium, 1713): plaktė ‘whip’ 25ᵇ (~ plaktę plakù), sutek te 44ᵃ, wogte 110ᵇ (~ vogtę pāvogė), gulte 131ᵃ (~ gultę gūla), patepte 252ᵇ (~ teptę tēpa).

69

La. east olūts has t from v; cf. āt-luša K̇vėd. from *āt-vuša instead of āt-uša (and that is how it is pronounced) with v from āt-vaša ‘La. offspring’; Telšiai parish vōnšōs, gen. vōnša > eastern Aukštaitian *vūšas, whence lūšas (Vadōkliai), alongside ūšas, pipe > eastern Aukštaitian *vuⁿzdys > luⁿzdys Dus. Kamajai.

Tekė-ti: tekė-stos, ėsty ‘выход замуж’ (marry, of a woman) in Linkmenys = ženyties < женu-ться: žený-stos in Linkmenys ‘marriage’ = German bevanō-n ‘to serve’: bevanō-sta neuter ‘service’. Fick III⁴ 177.

This autumn I was in the Ciskoda parish (12 versts from Rēzekne — Режица) studying the Lithuanian colony. While studying the Lithuanian dialect, I also jotted down a few things in Latvian. Here are a few examples (I cannot vouch that I heard them correctly): màrga ‘a mature girl, from her first menstruation’, mirdzinuot ‘to drizzle’, leitins kai mygla slacinōj, patmal’as ‘mill’, place-names ~ Cyskods, Vil’āni, nu, Gal’āni, -nu, Varkl’ōni, -nu, Rykova, Bykova, Drycāni, nu, Malnova, Lubuons—lake, dīndōrzs ~ klāvs, V̄orkova, Raznas uzars, Maltas river, ai vanagi, vanadžēni! vocative (in a song), pymu, pylñs, zeml’i ‘down’, myl’ti, aviša ‘raspberry’, sēni laiki, but vacs cīlvāks; malny muni kumelēni, kai yudēna vogul’ēni; gen. pl. žūsu, ziv-u; syvāns, but šīvenēnš, šīvenu jīmti, pyrmu dīn l’audavēna, caturdīn, sasdīn, ovu boltas kuojas, lyku lyku gal’vēnā (sc. a little wreath), meita dzaltuohim motim, munu vainūku atpakal’ atnas. Where did you put the wreath? I put it… in my lap. I will not go among the people without a crown. Don’t play, without scratching. garvilce ‘rainbow’, carmuška (=in Rēzekne, čermyuška), dzil’tēne, kuormi ‘trobesy, building’ from Prussian. *kārmi < Proto-Russian *хорми ‘mansions’.

I was in Ciskoda for only four days. I had no proper opportunity to learn the accent. At the time I was more concerned with Lithuanian than Latvian dialects. Both the Latvian and Lithuanian dialects seemed very interesting to me; I plan to go again another time.

More about the Curonians.

The most important features of the Curonian language are found in the word *Santace = Lithuanian san-takys: ‘a stream where two streams meet’ Sentatze, where two streams join, 1422. Bielenstein, Grenzen, 452.

Curonian bruk Bente-pürge 1422, ibid. (eyn brok Sate-purve 1439, ibid., 455.

Curonian ‘stream by the bridge’ Grynde 1422, ibid., 452: Lithuanian grind-inyš ~ Latvian Zemgalian grīda.

Curonian Kandamer Vytzen sture, ibid., 453.

Curonian Cantimas 1387, ibid., 450; 1439, ibid., 455 ~ Cantyn 1290, ibid., 440.

Curonian Mames-pintes ibid., 477 ~ Prussian pintis ‘road’?

Curonian Svente farmstead names among Zabels and among Matkuls (both not far from the lower Abava), ibid., 344.

Curonian Dzintere ibid., 282²¹⁸: Lithuanian Gintara, river.

{ stream Zynthere 480

Curonian Wanderen 1253, ibid., 233.

Curonian Stembre 1386, ibid., 450: Latvian Zemgalian stiēbrs ‘stalk’ ~ Samogitian stembrai ‘a kind of reed’ in Mosėdis.

Curonian. Venzava l. c. 195. 198.

70

Curonian Alš-vanga (Biel. Grenzen 206) derives from *Alšu vanga: *Alšu from *alš(ń)juo “alder”. Alšu: *alsnis (> al’ksnis) = krašu Biel. Lett. Spr. II 4: krãsni-, cf. g. pl. zvaĩžu l. c. II 32 ~ *zvaizne.

Curonian Alšu muiža = Aĩkšnių (estate).

Curonian levanga will be lev-vanga from ievų vanga.

What do you think, is there anything Celtic in the land of the Balts, as A. Shakhmatov claims (AfSPh. XXXIII 51 ff.)?

I find nothing Celtic among the other names, nor among the following ones: Nė-munas l. c. 77 (cf. оз. Муна ~ застенок По-муне near Lake Dykšnykšcio and the Dysna River, Švenčionys County. Cf. He-vapda: Vãrd-uvos river near Alsėdžiai = не-веда: Вед-epa Prussian Wedera Wedir lacus Ness. = Ne-vėžis: Vėžio river, Kvěd. p. Не-коча > >Samogitian Nãkačia (here they say na-žinau): Ком-pа = Не-мокшита lake: Makštýs Lake), Лучеса l. c. 79 = Laukėsà tributary of the Daugava near Gryva, Latvian Riga l. c. 81 = riņga ~ Riņģys stream near Ušpaliai.

Perhaps I shall remain in Mosėdis a little longer. Please write to me at the same address. I wish you all the best for the coming New Year.

I remain, with respect, your Kazimieras Būga.

[6]

Mosėdis. January 8 (21), 1912.

Dear Friend!

The kl’ėvas of the Dusetos–Kupiškis dialect speakers is intelligible to me only from Proto-Indo-European. kl’avas; cf. Latvian klava. From Proto-Indo-European *klėvas one would expect *klėvas in Dusetos, and *klāvas in Kupiškis. I cannot say how they pronounce in Dusetos Kup. lėpinu, (iš)lėpęs, however (I no longer remember).

The people of Dusetos say: lėkia, slėgia, plėšia, tęsa, lėidas, plėkšta, klėtys neuter plural, glėbja genitive singular, lãidžiu, plãisėt, but lėpšys ‘a kind of mushroom’, tãvo aũsys nu-l’ėpę (= ‘bent over’) like pigs ~ l’ėpštų l’ėpti, l’eipstų juokù ‘I burst into laughter, I can’t stop laughing’ (cf. leipti in Jušk.’s dictionary I 422ᵇ, 500ᵇ), bl’édis, -ies “damage” (cf. Лхр. 208₁₂), gl’eizoti “to crack” in Linkmenys.

Bl’ėdis, apparently, is a word of Latvian origin.

For the alternation of intonation (vãrna: vãrnas), one may also compare the following sequence of examples:

Prussian ~ fem. Prussian woman Kved. Rietavas

neighbor ~ fem. female neighbor “ ” ”, cf. the Samogitian doctor ~ fem. female doctor “доктор (женщина)” in the Viekšniai parish.

The Telšiai form ãštõnc, if derived from *aštountas, would have developed later, like the genitive plural vỹru (with a close u) < *vyrų < *vyrun < *vyruon. Útõnc should be compared (but is not identical!) with the Telšiai form vėnc (genitive singular vėina) ‘one’ [ė is not nasal, i.e. not *vėnc!]. The Telšiai Samogitians say ãštõunė.

71

Telš. Samogitian acc. pl. tūs vyrus, like Tamsta, I now explain from *tuns vỹrųns [cf. gen. pl. of the Mosėdis dialect: tũ vỹru alongside nom. du. tõu-do vỹrō] < *tións vỹruons. Telš. Samogitian acc. pl. pėktũs vỹrus: of the Raseiniai Samogitians (Kved. Riet.) pėktūns-õs [the Raseiniai Samogitians no longer distinguish the short vowels u ~ õ, i ~ ė] vỹrõs.

Pėktas mergãs Kv.: pėktūns-es f. Kv. = pėktõs vỹrõs Kv.: pėktūns-õs m. Kv. Telšiai Samogitians say only pėktũos-es (múotėris) < piktós-ias; cf. Dus. piktãs-ias. The Telšiai Samogitians' acc. pl. pėktũsius would derive from *pėktūnsius; cf. Telš. Samogitian skūšti.

Thus it appears that words with an ō-stem at one time had two possible endings in the acc. pl.: -õs and -ōns.

The Telš. Samogitian acc. pl. tūs vỹrus and gen. pl. vỹru show that aštõnc cannot derive from *aštuontas. From *aštōunc (which would be a very recent coinage) aštõnc: *aštōunc > *aštõnc could have developed [cf. kąs vėn from *vėin < *veinõ ntr. (cf. Telš. Samogitian neuter trūmpō) < *viė-nã] > aštõnc; cf. dat. du. dõm vỹrōm < *dōum.

Lithuanian kuodėlis, it seems to me, is a word of Russian origin (cf. Low German põukā < Russian пух. For uo cf. Latvian duoma, kuoma etc.). Mikkola considers the Finnish word kuontalo Russian, but it could also be Baltic. If kuontalo were a Baltic word, one could infer from it that in Proto-Baltic ān + Cons. had not yet merged with ōn + Cons.

Balt. kānd- > f. kuont-alo: Daukantas's fem. acc. sg. aną tą [beside masc. anou, tou. Prasmą lotinų kalbõs... Petrapilie 1837 p. 19. In all his writings Daukantas distinguishes tą, i.e. tā, from tou] < *anān *tān: tūs < *tōns: vilkui: ž. vỹru: pùlti.

With all this in mind, deriving -ui- from -uoi- < i.-eur. -ōi- is also not impossible. Lith. lūina РФВ. LXVI 225 could derive from Indo-European lōinā: Celtic leino-; cf. what Bezzenb. says in K Z. XLIV 309 (I was able to read it at the Vilnius Scientific Society).

But could slūnkius mean “der Schleicher” (‘the creeper’), from *slōnk-? Slūnkius < *slōnk-: slankius < slōnk-: slenkù = Latvian kluinis: Latvian klainis cf. BW. 29840: kleinis ‘bow-legged’ BW. 11308 var. = Latvian krūilis: *krailis: kreilis РФВ. LXVI 222 = luitas ibid. 226: laitas: liėti ibid. 227 = muitā: maitā: meit-kaulis = puĩšos: paĩšos: piẽšos. = tūlp-in-ti Jušk. ž. I 554a: talp-inti: telpù = į-tumpas (in Kuršaitis's parentheses): tampau: tęmpti = Latvian tumsa: Lithuanian tamsa: témti.

The question about -ōi- is still not clear to me. From Indo-European ōu, it seems, one should expect ū:: ōu > *uou > uu (cf. uoi > uū) > ū; cf. pa-laid-uonas (cf. Лхр. 70. Širvydas's Dict. 9, 165, 277, 359 – palayduo) and pa-laidūnas < *-ōunos? Palaid-uō < *palaiduo[u]n, but gen. sg. *palaidōn-ės > *palaidūn-?

The Livonian mändreks, kill, loiga, koidas cannot be of Samogitian origin, since according to Bielenstein (Die Grenzen) the Livonians were not neighbors of the Samogitians: Livonian place names do not extend as far as Gramžda. Livonian loiga, koidas must be Curonian words. The Livonian oi would derive from Curonian ei (as in Telšiai Samogitian).

Could the Curonians perhaps be Latvianized Samogitians? The Livonian kill = Latvian dzilna would indicate that the Livonians had come into contact with the Curonians—Samogitians even before Latvianization.

All eastern Lithuanians know only kiškį (Salõčiai, Panevėžys County; Dus.; Širvydas). The Telšiai and Raseiniai Samogitians, and the western Aukštaitians (Tilžė, S. Naumiestis) say zuikis. Zuikis could be a Curonian word, *zuojekias. From *zuojekio, Samogitian must have developed *zuojikys: unstressed e yields i ~ Telš. ōudėgā = Rasein. ūdėgā etc. from uodigā <

72

tail. Trejetas uodegà. Trėjėtas > žem. *trėjits > treits > trėits > trẽc “three” (Mos. Salantai). *Žuõjikys > *žõujikys > *žõu[j]ikys > žõikys (as pronounced by the Mōsėdiškiai and Salantai Samogitians) > Tilžė zuikis.

In Tverečius, apparently, the instrumental singular is sù širdziu, as some also say in Dusetos — širdžiu.

The Telšiai Samogitian accusative singular of vỹrs is vỹra, but the feminine instrumental singular is sõ bõrnõ. For this distinction, cf. Dusetos vỹrų, but sù bũrnų; Suvalkai Naumiestis and Virbalis (Вержболово) — vỹrą, sù bùrną.

Would it be possible to obtain Dzimtenes Vēstnesis, where you wrote about Shakhmatov’s raised forms?

Aistere also has Aistar-? cf. Aistar-nieki BW. 12684 (Rāvā).

Tuos-mare BW. 9593 (on the lakeshore) is related in its first part to the river name Tuos-ele (a tributary of the Bārta near Nicās. Konverzacijos Wārdnica 279). In its second part Tuos-mare is comparable with the Latvian town name Palc-mare (Konv. vārdn. 667), situated on the bank of the Palce (a tributary of the Gauja): Lithuanian Žiž[ma]-māriai, a town in Trakai County; Žižmariškiai, a village in Vabalninkas parish ~ cf. Žižmà, 1) a river in Kuršėnai parish, Tryškiai district; 2) a tributary of the Gauja in Dieveniškės parish, Vilnius Governorate.

-mare: Marra, a tributary of the Ryva, which flows into the Baltic Sea at Labraggen. Bielenstein, Grenzen 480: Lithuanian XVI-century stream Мор-иха (Ревизия душь...50, 54) ~ from the tract Поставяли (= Pastovėliai), a mile to Lake Моря 1. p. 54 (Seinai County, Suwałki Governorate): Marūge, a stream near the town of Vidz — Vidžiai: Pa-mar-ēika, a river near Zasečiai, in Dyatlovo volost, Slonim district; Prussian Pas-mar, a river near Kreuzburg (Nesselmann, Thes.):? Old Indian mārayati ‘submerges something, throws into water’.

It is unclear whether -mare is mārē ‘sea’.

Curonian Vec- Sumber—a farmstead in the Dondangen district (Biel. Grenz. 288): sūbr-īti! BW. 30493 in Skrunda, from the Semigallian dialect.

Curonian And-ume near Stende ~ Talsi (Konverz. vārdnīca 113, 1144):? Slavic qda.

{cf. venter-kājis BW. 14517, 9 in Kuldīga

Venteru farmsteads, in a deep river valley (near Aizupe, in Talsi district). Konverz. vārdn. 47: vēnteris masc. Mosėdis, Dusetos = Russian вѧтерь.

Rindz-ele in the vicinity of Engure. Konverz. vārdn. 846: Samogitian Ringà < Curonian *Ringa = Semigallian Rīga.

Curonian Alšu muiza, Alš(u) vanga; cf. alkšu rikste BW. 9778, 1 variant, Dundaga: lagžu krūminā BW. 29316.

I do not know when Volter will publish his studies of Slonim (they were supposed to appear in Изв. И.А.Н. at the beginning of 1912).

When giving literary-language forms in scholarly articles, I would also like to use literary orthography (as you did in Слав.-бал. эт.): for example, vỹras, eĩsiu, žõdžio. When giving dialect forms, scholarly orthography should be used; cf.: Telšiai Samogitian dative singular vīrou, first-person singular future dōusō, Telšiai Samogitian būužava, Tverečius žãgìs, žìdel’is.

In examples of written language, I would recommend following the ordinary spelling of contemporary written texts, that is, writing: vỹras, vērksiu, mēdžio.

73

The word paveikslas did not arise by this route: pa-veid + tlas (cf. *ár-tlas > or-klas) > *pa-veistlas > *paveikslas [not *paveišklas, because when it was taut-iškas with š, paveikslas still retained a dental t before s] > paveikslas.

In the next issue of РфБ, my “Slavic-Baltic Etymologies” will appear.

Best wishes,

Yours, K. Būga.

P.S. I read Bezzemb.'s review while I was in Vilnius.

[7]

Mósėdis, January (I) 30, 1912.

Dear Friend!

From Indo-European ōu I would expect ū in Lithuanian: ōu > *uou (cf. -uoi > -ui) > *uu > ū. Alongside Daukša’s nom. sg. žmuo, he has acc. sg. žmūnį, which I would like to interpret as follows: žmuo < *žmō < *žmō(n) from *žmōun, cf. ákmūo ~ ἄκμων, but žmūnį from *žmōun-m. Širvydas’s širšuo should likewise be understood together with the Telšiai Samogitian širšūns < širšūnas.

Latvian pērk-uon-s: perkūnas

did(u)omenė Dusetos dialect: didūmenė Salantai

sūodys plural ‘rye’: sūdys Telšiai Samogitian

grūodas: Slavic gruda < *grōuda?

šuokoti intensive form: šuk-snis masculine ‘leap’

liūbas: Slavic lubъ < *loub(h)os: liūbos

ὠμός: ūmas ‘raw’: um-ēdē ‘russula’

miegūostas (cf. Russian зуб-аст: miegūstas

snaigulė: Telšiai Samogitian snaigūlė.

The Telšiai Samogitian aštōnc ‘eighth’ is unclear to me: from *aštōuntas one would expect *aštūnc; cf. alongside pōudōnktis ‘pot lid’ we also have pūndōnktis. The latter form derives from *pōuddōnktis (dissimilation of two d’s. Cf. the form andiot < *adduoti ‘to give’ in the village of Zasėčiai (Slānimo district)) > *pōundōnktis (cf. kailiniai > Dus. kalniai < *kailniai) > *pōndōnktis > *pūndōnktis > pūndōnktis. Pūndōnktis is more likely from *pōndānktis (when an + Con. had not yet changed into ōn + Con).

Pērkuons ~ perkūnas may also be interpreted differently; cf. χελῶνη ~ χελῶνα.

Telšiai Samogitian aštōnc, it seems to me, can only be derived from *aštāntas, which would indicate that ōn changes not into un, but into an. *Aštāntas, together with the instrumental plural ending -ais = Greek -οις and Sanskrit aiś, would indicate that ō was shortened in closed syllables in Proto-Baltic. The failure of the diphthongs -ōi (vilkui) and -ōn (žąs-ų ἔχην-ῶν) to shorten can be understood through sandhi ([in positione] before the vowel of the following word). On this account, the Samogitian accusative plural -us: telš. tūs vėlkuos (beside dô vėlkō!) remains unexplained. Telšiai Samogitian accusative plural tūs from *tūns; for the close quality of u compare pūdamenc, gen. sg. pūdamēnta ‘foundation’ alongside pūndamēnc, gen. sg. pūndamēnta.

74

Tùs < [*tųs?] < *tūns < *tūons (cf. Greek τούς) = gen. pl. vārṭu < *vārṭu < *vārṭun < *vārṭuon.

The Aukštaitian tuõs would be explainable from *tuõns < *tuõns only if we assumed that in the Aukštaitian dialect “vowel+n+s” changed into “nasal vowel+s” earlier than in Samogitian, and that those diphthongs began to change into “nasal vowels” before shortening uon into un. The -uont- had still not shortened to -unt- when the Aukštaitians had already begun pronouncing *tuõš.

That -uøs could have changed into -uos seems beyond doubt.

That in the Aukštaitian dialect an+s, en+s changed long ago into ąs, ęs can be seen from the Dzūkians of Trakai and Seinai districts: alongside ranką, renkų they say žūsis, tįsų < žąsis, tęsiu.

In Samogitian dialects *žansis changed into *žęsís much later than in Aukštaitian.

The Samogitians of Žvingiai parish (Raseiniai district) still say žansės today (with nasal a+n). Kvėd. žąnsės: Kuliiai, Salantai žōsės: Mosėdis, Darbėnai žōusės.

The Telšiai Samogitian accusative plural tùs = and Aukštaitian tuõs show that o in closed syllables was not shortened to a in Proto-Baltic (but how should *aštañtą be understood?).

Lithuanian lėkas apparently derives from Indo-European *leikʷont-s (cf. Sanskrit bhāran ~ *bhāras t-), and not from *leikʷōnts (cf. Greek φέρων[t], Slavic bery).

Rūšis: rāisas = Samogitian rūinis ‘speckled one’ (Jaunius): Samogitian rāinas = klūkis: *klaikas > pa-klaikti (cf. sveikas > pa-sveikti).

Latvian krūlis: kreil(a)-s = Latvian kluinis: kleins.

Gruinys: grýnas.

Here we have a whole series of substantives in ui, derived from adjectives. Could ui here be from Indo-European ōi? It is difficult to understand ui in this case as coming from anything else.

Just as alongside rāisas we have rūšis, so alongside sēnas we ought to have *sānis, and alongside šēmas – *šūomis. Without *sānis, *šūomis, there would be no need to explain the word rūšis from *rōišīs either. If rūšis is not *rōišis, then what is it? Latvian krūlis and kluinis show that words for varieties of rūšis descend from the common Latvian-Lithuanian proto-language.

Sēnis, šēmis, it seems, cannot be recent formations replacing *sānis, *šūomis.

The question of the diphthong ui is very complicated and, in my opinion, still unresolved today.

Zuikis could be Curonian *zuicis. This word would have been borrowed back when the Curonians still said *zuikis; cf. Curonian Rucava =, Samogitian Rōkava (ō of intermediate length) < Rūkava. Curonian *zuicis from *zuoicis (with epenthetic i) < *zuõčis; Latvian zakis instead of *zacis with dialectal k (of the Ventians?? Semigallians??), or with k from *zak(a)s? *Zuočis: *zacis would have different vowel grades (Ablaut!). The relationship to заяц, Russian зайка, заенька, заинька is unclear.

Заяц < *Žo-j-en-ko-s, заень-[ька] < *žo-j-en-[ikā] from Indo-European *ǵ(h)ō[u?]-jo-ar ǵ(h)ōj-o- < ǵ(h)ō[i]-.

Curonian *zuõ-c-is ~ Latvian za-k-is: Slavic za-j-en- = Lithuanian ō-yum Latvian uo-l-a ‘egg’: Slavic ja-j-e.

1) The Lithuanian-Latvian diphthong ie yields ei in Finnish: heinā < ‘hay’, seinā < ‘wall’.

Baltic ei (from which Lithuanian-Latvian ie derives) yields Finnish ai: taivas < *deivas, paimen < *peimen-, vāivel['] from *vaives < *veives-[ā] (whence Lithuanian vievesà, -õs, accusative vievesą. Kv. Mōsedis).

75

2) Lithuanian — Latvian. uo = Finnish uu: juutas < juõds, luuta < šluota.

Baltic ō = Finnish ou: routa < *grōdas “grúodas.”

From this we see two periods in Baltic-Finnic relations: 1) the earlier period—the Baltic period; 2) the later period—the Latvian-Lithuanian (or Lithuanian alone?) period. Finnish morsian “bride,” lohi “salmon,” toe “Latvian tacis” would be words from the earlier Baltic period, when short a was still distinguished from short o. The Baltic origin of these words is beyond doubt. Finnish olut may be Germanic or Baltic (cf. Prussian alu from *alut). Finnic o in the words mentioned can only be explained from Baltic ō: from a we would also have Finnic a. Morsian etc. would be older than the words of the Baltic period taivas and routa. morsian etc. could also be of the same age as taivas, routa, if we accept that in the endings (without stress?) ō changed to ā earlier than in other environments (under stress, or also without stress?).

The question of short o is very important. We should inquire here of the Finnish scholars what they say about morsian etc. It seems to me that morsian cannot be explained otherwise.

Is Finnish kuontalo a Baltic word (Latvian kuodel’a, if it is not a Russian word), or a Slavic word? I cannot understand the Finnic -uont- as deriving from Slavic *kođĕlь.

If Lithuanian dažgys derives from *dagijas (cf. = Livonian tagijas), then why is galvijas not *galvys? (kalavijas etc.). Zemaitian and eastern Aukštaitian kirvys could also be understood as deriving from *kirvij(a)s (with a far from the stress!). If Livonian tagijas “thistle” cannot be understood in any other way than as *dagijas, then dagy̌s will have to be read as deriving from *dagijas. Are there more Finnish forms in -ija(s) from Baltic -is? I do not have Thomsen’s work with me at present, so I cannot look for more words in -ija(s).

I wish you all the best. Yours respectfully, Kaz. Būga.

P.S. I will send you my new address from Vilnius, where I will be on February 3 and 4. Please write to me in Vilnius at this address:

Leatherworkers’ Street, No. 5, editorial office of the newspaper “Viltis.”

[8]

Pažiegė estate, Dusetos district, March 17 (30), 1912.

Dear Friend!

Congratulations on successfully defending your doctoral dissertation! I hope the ministry will soon approve your appointment as professor!

Baudouin is an eccentric and incomprehensible man. Although he is my teacher, there is much in his character and conduct that I dislike. B. has deteriorated most in recent years. The objections that Shakhmatov made must be printed. They say that Sh. has some additions to your paper. I have not had a chance to speak with Sh. about this matter.

Your article about -ui- will be awaited not only by me, but also by European scholars. The question of the diphthong -ui- remains unexplained for European linguists as well. I gladly give you permission to publish my trifles on the Samogitian dialect of Telšiai

76

The accusative plural ending is -us. At the end of summer I, too, would like to describe the Telšiai Samogitian “Auslaut” (-ė: -i, -ō: -u etc.).

The hearing of the diphthong -ui- has now become clear to me as well. Among other things, -ui- may also derive from i.-eur. -ōi-. Your reasoning that -ui- can arise only from stressed -ōi- seems correct to me. Rūnis beside rainas, rūišis beside raišas, kluikis beside klaik-šas ‘fatulus’ fully confirm that reasoning.

lui-ta-s ‘zamazka’ (with fixed stress. Cf. grąž-tai, kál-tai, saī-tai, skāb-tai etc.: Gr. φόρ-το-ς, πλοῦ-το-ς) from *lōitos: praet. lėjau (cf. Dus.: šlėjau, lėjau).

We have far fewer instances of -ui- than of -ai-; but there is no reason to be surprised by this. -ai- instead of the expected -ui- must be a Lithuanian-Latvian innovation, just like -ā- instead of -uo-; cf. sėdu, bėgu, slėgiu, but sodinù, boginù, sloginù instead of *suodinù (cf. slūogas, rūozas < režiu. plūoštas = Slav. пласть < plėšiu etc.).

Piẽšà: paišà: puišà Ds.¹ = slìnkìus (Kuršaitis’s slinkis in parentheses. Cf. mižnius: mižnìs Dus.): slankius: slùnkìus.

Lūiba PФB. LXVI 224: láibas = lūina: leìctas) = šlūina (Latv. slūns ‘wanderer’ Rīg. r. kr. XV 136): Latv. slaists loc. cit. 94 ~ šlaistāusi ~ šliēti. Words of this type have fixed stress on the root; cf. klėiva: klėivas, kváila ‘fool’: kváilas, kváisa ‘fool’: kváisas.

Latv. gaiba ‘that which perishes’ could be an innovation replacing *guība: adject. *gaibs = Lith. *gaibùs (geibti: *gaibùs = teikti: taikùs = nėšti: našùs etc.). Latv. (viñs g luži sa-)guībis (Rīg. r. kr. XV 77: guībt) would be an innovation replacing *sa-gaibis from *guība (cf. Lith. gvāiba).

Alongside mūistau, as noted by Koss. Litu. 48ª, the eastern Highlanders also say māistau ‘I chew’. But is the root not mēi-? (cf. lai-stau ~ lie-ti, sai-stau ~ sie-ti etc.)

uīdiju PФB. LXVI 231 is not derived from *vuōidijuo: vaidas.

Latv. plēst: pluosīt = Lith. *mieti?: mūistau.

Maitā: muitā, gen. sg. muitos Ds. < root mei- = laitas: luitas ‘zamazka’.

Proto-Indo-European ōu in Lithuanian-Latvian languages yields either uo or au: ōu > uo, but unstressed ōu > au.

kúopa = Slav. kupa

grúodas = “ gruda

liùobas = “ lubъ.

Latv. virsuone: virsaune, pērkuons: pērkauns.

Lithuanian–Latvian uo: au from Proto-Indo-European ōu has a complete analogy with Lithuanian–Latvian ie: ei from Proto-Indo-European ēi. Thus it follows that Lithuanian–Latvian uo may have a twofold origin: from Proto-Indo-European ō and ōu.

Nom. pl. lýguos in Tverečius: lýgūs is identical to kárties in Tverečius, Adūtiškis: kártis and, perhaps, to

didùomenė: didūmenė and

lapiẽnė: Samogitian lapinė.

Good day! Yours, Kaz. Būgà.

¹ Puišà is a neologism in place of paišà, from the adjective puišinas < Proto-Indo-European pōfkino-s: kìrtis (stress) was formerly fixed (cf. Russian красный, but краса).

77

Post, St. Dusėtai, Kovno Governorate, Zastenok Pažėte (Pažiegė). According to what Kārlis Petersons says in Aūčė, suikis, ruize are pig names – клички свиней suikis ‘half-pig’, sivėns ‘piglet’.

Aūce.

[9]

Pažiegė (Dusetos district), July 24, 1913.

Dear Friend!

I do not know whether this little letter of mine will find you by the seaside. I had been meaning to write for a long time, but I myself do not know how it happened. I spent 24 days in Vilnius. I spent 15 days visiting in Suvalkų Governorate (Naumiestis parish). In Vilnius I found quite a lot of material for a Lithuanian-language dictionary. I did not manage to use all the material during those days. I will have to go again at a more convenient time. Manuscript material for the dictionary is to be found not only in Vilnius, but also in Kaunas, Sejny, and other places. I found two more dictionary redactions by Juškevičius, not as complete as the printed one.

I am now studying Sanskrit. In the autumn I plan to take examinations in comparative grammar and Sanskrit. In the spring I will leave Latin and Greek. To study Sanskrit, I use G. Bühler’s work “Leitfaden für den Elementarcursus des Sanskrit” and Wackernagel.

I received your “Miszellen” from “Lietuvių Tauta” from Basanavičius, so please do not send them to me again. While in the countryside, I have not yet had a chance to see anything new published in the science of languages.

I will stay in Pažiegė until 10–15 VIII. From 15 VIII I will live in Petersburg at the same place, i.e. M. Дворянская № 7.

Now I will answer this and that in your letter, which I received while I was still in Vilnius.

1. That Lithuanian srovė ‘current’ should have arisen from an older form strovė is hardly something anyone can believe, since linguistic facts contradict it. The people of Dusetos say stryptie, strėnos, strūkt’!, but srėbtie, sriegas, plural sriegaī, srauja ūpė; the people of Kvedarna say strėnos, strypailiōti, strūgas, but srėbti, srauni ūpė. The Dzūkai of the Trakai and Seinai districts have str- everywhere in place of sr- and šr- and -sr-, -šr-; cf. srėbė, auštrà.

2. It also seems to me that axe is not a recent coinage. (cf. axe, broad).

Lengvì, lengvìos: lengvūs instead of *lengùs (with v lengvì feminine). Cf. erdvì feminine, > erdvūs masculine (*erdūs instead). Although lengvì, erdvì may have v by analogy with the related forms lengvà, erdvà (Samogitian lėngvas, eŕdvas).

I know of no such facts as would show kirvio ‘of an axe’ not to be a recent formation.

3. In all the dialects known to me, there are words alongside those with sl- and those with skl-. Alongside slidūs, sliekas, slėgiu, slapūkas, slāstai, the Dusetos dialect also has skl’ėdinas ‘full as an eye’, pà-sklìtė, 3rd person past ‘spilled out’, in-sklimbai, adverb ‘obliquely’, and skliaūtai.

In Indo-European, skl- does not become sl-. in Lithuanian dialects.

4. I no longer believe that žaltys and žalktys (Dusetos, Dysna: not *žalktýś!) arose from two separate forms. Žaltys = Latvian záltis derives from *žalk-tyś: -kt- phonetically changes to -t-. -kt- can remain only where the connection with the root from which the word was formed has not yet disappeared.

78

Smagītis, not *smaitis, because of smeigiū, smāgas; rāktas, not *rātas, because of rakinu, āt-rakas. In some dialects the connection with the root may have been severed earlier than in others. This explains High Lithuanian drūtas alongside Low Lithuanian drūktas. But the root itself, drūk- or drūg-, must have disappeared. We have dáiktą (not *dáitą) because until quite recently we still had *dáika (cf. Latvian daiki, daiceklis). Prussian aukta- ‘high’ yielded the form auta- in the Sambian dialect (cf. Gerullis). The ancient Lithuanian name *Daugi-tāras is now pronounced only as Dautāras (*Dauktāras). Cf. Daug-vilā: Vaiš-tāras.

5. The forms eisina, bēgsina, etc., in the Samogitian dialect of Teišiškiai may have arisen phonetically from eisena < *eisianga. Cf. Telšiai lētina, uodigà, kūpita, etc. Samogitian -sina < -sena < -sjanā is equivalent through sound change to the Telšiai Samogitian word mēdiga < *mēdega (cf. Samogitian uodigà) < *mēdjęga < Proto-Samogitian *mēdjugā = High Lithuanian mēdžiaga.

6. The Latvian diminutives naksniņa, sirsniņa, etc., with -sn- are also incomprehensible to me.

7. Balžiena in Dusetos (Samogitian balžena in Kvėdarna), it seems to me, is not equivalent in its suffix to Latvian -ēns (edīens, metīens). Balž-iena (I have not heard the form balžienas in spoken language) is a derivative of *balžā-; cf. kiaul-iena in Dusetos (Samogitian kiaul-ēna in Kvėdarna, Mosėdis) || kiaulē.

8. Leskien’s molūjas is not a genuine word form: Juškevičius has, as far as I remember (помню), accusative singular mólūji pūdą (from molus, moli?). Latvian māluojis is equivalent in its ending to the place-name Kalnuojai.

A glossary is to be published with Volter’s chrestomathy; its manuscript is already complete (prepared by Juozas Balčikonis, a graduate of the Slavic–Russian department of the University of St Petersburg).

I will add a few more small points:

9. Lithuanian moteriškas ‘woman’: παῖ-ς-χος ‘lad’.

10. Laiškūs, but laikš-ūs. Aiškus: aikš-tē.

11. Image < *paveikslas (-skl- > -ksl-) < *paveistklas < *pa-vēizd-klas (: veizdēti: ār-klas, ir-klas etc.).

12. Boginti a new formation, to plant, to depress: I run, I sit, I press, but the ancient form survives in the forms: bēga pa-šūokom(is) Dus. “while jumping about,” rūož-tas “line, furrow” || to cut, to split || to tear (for ablaut cf. Greek φέρω-τος || φέρω, πλοῦ-τος || πλέω etc., Lithuanian grāž-ta-s || grēžti, sai-tas || siēti etc.). Cf. Latvian pluosīt: plēst.

Jumping: pa-šūokom(is), pressing: slūogas shows that slogā, Samogitian sodā “village,” žolē etc. are new formations.

13. In place of Russian u, Lithuanians, like Latvians, have uo: kruōpos, kruōpų < крупа, kuo dėlis < кудель *къделъ, puōkas < пух.

14. In place of Russian i we have ie in the word pielā, pielōs, pielą (Dus.) < nunā.

15. Prussian klanti- “to curse,” from Polish, is incomprehensible. There is no klanti(tvei) denominative from *klan-tā- “oath” < Indo-European *klōn-tā (cf. χολ-τη || χεῖμα, Lithuanian lak-tā, g. lāktos || lekū, load, g. nāš-tos || nešti etc.): Slavic klьno, klēti < klen-?

16. Maitā “carrion”: meit-ēlis “penis” || mitas || mitinti = German Aas “carrion”: to eat.

17. Tūmulas “lump, clod” (snow lumps are beating against the panes. Viltis 1912 no. 36 Jušk. dict. I 286ᵃ) Latin tumulus: Lithuanian tum-tas.

79

18. Mailius ‘tiny fish’ collect. ‘мелзота’ Dus.: Slavic mélz(k)o.

19. Fashion, fashion “мода” Svėdasai, Kūniškis: Slavic to carry “не́си́ти”. flowing, might flow “flowing, = выход замуж” Linkmenės: to marry “выходить замуж”. life, gen. life “жизнь” Dus.: alive. Pity “жалость, сожаление” Dus., An. šil.:? humanity, human nature etc.

Perhaps you are traveling around Latvia this summer on dialectological business?

Best wishes!

Kaz. Būga

Post Office, Dusetos, Kovno Governorate.

From Pažiegė to Kazimieras Būga

[10]

1. The Lithuanians of Trakai County, as far as I have been able to ascertain, do not change the diphthongs an, en into un, in, as the people of Dusetos, Anykščiai, and Tverečius do, etc.

In the parish of Oniškis (Ганушишки), for example, they say: anciulė ‘duckling’ Лхр. 390₂₀, krantų acc. sg. l. c. 390₂₂, mendr- l. c. 390₃₄₃₅₃₇ = Dusetos mindr-, nemieganciu acc. sg. fem. l. c. 391₃₈, danguĩ l. c. 394₈, peŕ langély l. c. 395₂₅, etc.

In the vicinity of Jieznas, Punia, Marcinkonys, and Merkinė (= Меречь), they also say dantis, rankà, dengiu, etc.

Dat. sing. žąsi (in Jieznas they pronounce žāši; cf. grąžtas < grąžtas), šuni, etc., I know from Jablonskis.

Jablonskis has heard a short (not half-long!) u in the dative singular šuni. In Trakai County, it seems, there are no half-long vowels at all—at least Jablonskis has not heard any in Punia, Valkininkai, or Jieznas.

2. I explained the endings of the Elbing glossary -ay (brookay 480 ‘trousers’, nom. du.) and -oy (clattay 292 ‘clogs’) by the accent: -ay from -aĩ (nom. du. fem.), and -oy from -aĩ (nom. pl. masc.). -aĩ in the glossary dialect must have changed into -ũĩ; cf. dalgių 546 = dalgis, corbis 306 = kàrbas.

In the glossary dialect, it is not always clear for what reason we have o instead of the expected a:

api-sorx 772 ‘the kingfisher’ < -šãrks?

bordus 101 ‘beard’ < bàrdus?

dongo 403 ‘Refe’ < dāngà (cf. Lithuanian dialect òmžinė ‘forever’ alongside āmžinas in Baisogala)

carua-woytis 416 [Trautm. 353 writes incorrectly: carya-ą] < karjà-vāitis: kariowieſch-czias marginal gloss to Kariobalfus by Bretkūnas (BGLSpr. 293) = kārio vieščias acc. pl. etc. etc.

3. In the Dusetos dialect word pētū ‘south’ has an open (half-long) e, before which p is neither hard nor soft! The accusative piētūs has an unreduced diphthong, thanks to Baranauskas’s “stress reflex.” Such forms as piētūs arise thanks to Jaunius’s law concerning “the mutual force of colliding stresses” (Понев. гов. I ч., стр. 97 сл.). Instead of piētūs vérda, people more often say piētus vérda; cf. liētus sūnųšė

80

Rugiūs. Accent retraction in the Dusetos dialect is also found across a pause; cf. ą’ras! ‘get out!’ (said to drive a dog away), alongside ãras, aik tãlin! < tolỹn, kur árklius vẽsma? = (answer:) vesmã Terpūravėn.

4. Gerãmui-jamui could have changed into gerãmujamui, from which, by further development, >*gerãm-jam(ui) > gerãmjam (dissimilatio duorum m) > gerãjam. The people of Dusetos say gerã-jam (dat. and loc. sg.), with a half-long a.

Dusetos dialect dat. sg. fem. gerãjei, with a half-long a.

5. The form poniẽ of Buivydžiai is entirely unfamiliar to my friend Kilys, who also comes from Buivydžiai; he knows only pã’ni (with retracted stress, but mixed). One must be very cautious with word forms taken from R. Gauthiot’s paper. Gauthiot was insufficiently prepared to investigate Lithuanian dialects. The following observations should be made about Mr. Gauthiot:

1) he has not quite learned to distinguish intonations: he writes šẽmas 105, ruñka 35, tãmsta tãms-ta 106 [NB, because that is what he finds in Kursaitis!!], whereas in Buivydžiai (Petras Kilys) only šẽmas, ruñka (gestossen!), tãmsta are said;

2) he cannot distinguish soft p from pj: he writes pãuju 101 (cf. Kurš. liet.—vok. žod. 308: piãuju!) instead of pjãuju;

3) he cannot distinguish e (the middle sound between e and i) from e: trẽjų gírnų 83 is not from trijų, but from trẽjų, because people say not *tris girnas, but trẽjas girnas;

4) he writes gardži megójā 1 sg. 84 instead of the spoken gardž’ē mẽgojo: ē ā ō are half-long;

5) he treats an unstressed half-long vowel as short, but as long when stressed: skãrsaẽi kẽl’ù 82₁₄ instead of skãrsaẽi kẽl’ù;

6) he writes alã’ 84, whereas the pronunciation is al’ã’;

7) he writes pịã’nunt 83₆, pịã’nunt 83₁₃, whereas people pronounce only pãẽnunt (elsewhere pẽnint: infinitive penėti ‘to feed’).

6. The people of Dusetos usually say apẽ ‘apiẽ’ (*apẽ I have not heard) [with an open e]. The Lithuanians of the northwestern Suvalkai region (Naũmiestis, Slavikai, Sintautai) say: ãpė.

Alongside ter̃p, the people of Dusetos use the noun terpẽ ‘промежуток’, terpẽklis ‘т. с.’

7. Baranovsky, Notes 29: čianãi tenãi means ‘здесь = čia’, „там = tē“; kuriãi katrãi are dative singular. Dative singular feminine kuriãi katrãi is also said by the people of Dusetos.

Alongside čiãnãi tỹnãi, the people of Dusetos also say č’enãi tinãi ~ čianãi tynãi and čenãi ti-nãi. In Ramygala (not far from Panevėžys), they say činãi tinãi and činãjos tinãjos [cf. jinãjos ‘jinãi = она’].

The people of Dusetos’ tinãi, as shown by the full length of the word ending and ã, has recently been shortened from a longer form. Tinãi < tinãjās instead of *tinãjās with a circumflex accent, from tinãi-gi.

That G-t read pịã’nunt with a diphthong can be inferred from vijã’no 82₃ = viã’no ‘one’;

8) He does not distinguish soft ś from soft š: ušj 107 instead of ũš (with u of medium length).

Gauthiot’s gen. sg. sūnũ’s 41 is his phonetic misinterpretation of sūnõs < *sūnõũs; compare: senõ 81 ‘earlier’, tol’ọ 82, 83 ‘farther’, sirgõ ‘I fell ill’ 103, aš bijãjā 83, megójā ‘I slept’. G-as marked sūnõs in his transcription as sūnõs (cf. tol’ọ 82, 83 = tol’õ 106) and, following the example tadõ 82 = ta’ tadù 83, made it sūnũs. This sūnũs, invented out of whole cloth, already appears in Trautm. 240.

81

Do not include the Buivyždžiai form *poniė, as a very doubtful form, in your writing. I hope to be in Buivyždžiai this summer and check everything.

I repeat that the form *klėk < klėk is entirely unfamiliar to Petras Kilius of Buivyždžiai; he only says klėk.

8. I compare the loc. sg. *širdí, *suní, together with Jaunius, with Vedic (RV.) vėdi (Grassmann, Wb. zum Rig-Veda 1353) and čamū (l. c. 437).

9. The 1605 Catechism’s mielup’ sunúp’ 58₁, ‘apud carum filium’, cannot have originated from accus. sing. + p(i), since then we would have *mielump’ *sunúmp’: a “nasal” vowel before pi becomes “vocalis + m”; compare also the 1605 Catechism’s sawimṕ 38₁₈,₁₉, sáwimp 59₃, szwintúmpium 55₁₆, músump’ 58₁₁, maldump 71₃; 72₁₅ darbump gierump 74₉.

Daukša’s žmogúp’ likewise cannot derive from the acc. sg. žmogų + pi. From the latter form, Daukša would have obtained *žmogump’, but never žmogúp’.

10. Vėlnjas may be explained from the earlier form *vél’ nas < vélinas; compare Russian вольно = Lithuanian vál nia adverb (Dusetos dialect) < *vál’na.

11. Daukša’s úgnip’ cannot have originated from the acc. sg. úgnį + pi (the result would then be: *úgnimp’).

12. From *akì (< i.-eur. *ogēi)+en one should expect not akijė, but *akę.

13. How the forms úgnip’, ligónip’, Wieszgpatip etc. should be read can be shown by the Lazdūnai, Dieveniškis, and Zasėčiai dialects.

14. In loc. sg. mergo-jė žėmėj-e akij-ė I did not wish to seek gr. ėv for the sole reason that it corresponds to Latvian iē- =, Prussian ēn-, with an Indo-European circumflex.

15. Lithuanian loc. sg. laiškė, sakramentė is not to be distinguished from Prussian en… laisken III Cat. 63₁₆ and en Sacramenten 49₁₈,₂₉.

16. The Dusetiškiai túmsus ‘dark’ derives from tám sus, as shown by the Seinai parish forms tám sus, rankū acc. sg., pėmpį acc. sg., etc.

17. Samogitian dat. sg. úgnie, ávie, etc., may have the archaic -ie even without the influence of loc. sg. *ugnie[je] (which the Samogitians do not know at all!).

That Samogitian ѵ -ie does not turn into -i may be shown by the 3rd person permissive.

tavaguōji ~ tavaguōjėi

tavažiūji ~ tavažiūjėi

< te-vagōjie, te-važiūjė

Samogitian talingūji ~ telingōvėi: tanešī ~ tānešēi = miēnū ~ miēnōū: šū ~ šōū

But Samogitian nėšti: nėštīs ~ nėšteis

{gen. sg. kartis: širdīs ~ širdėis

18. Prussian ioūs is = la. jūs < in Indo-European. *jūs. The derivation of ioūs from *jūs is my oversight.

19. Daukša’s bū, in the 1605 Catechism búo (1 pl. búome), can hardly derive from *bū(v)o. Alongside būti: búvo > one would also have to expect bū from Daukša, as well as pūti: pūvo > *pū, žūti: žūvo > *žū., etc.

Búo-me (with this accentuation in the 1605 Catechism): bú-kla = German bō-pla neut. bō-pōn fem.: German bū-pō fem. Fick III⁴ 272.

20. Prussian pirmois masc., pirmoi fem. are probably derived from *pirma-(j)is:[*pirma-(j)i (Ulmann 37₃₁); compare Old Lithuanian geraji, szwentaji. Aist. stud. I p. 13.

82

21. In Aist. stud. I 99* I explained the form nom. pl. žaislai as a neologism derived from acc. pl. žāslus (pronounced thus in Slavikai, on the Prussian border, Suv. g., alongside nom. žaislai), by analogy with Jókūbaitis ~ Nau jokátis (both forms in Slavikai). This explanation, I think, can be accepted.

Ąžuols instead of ąžuols with ai, from nom. pl. ąžuolai, a neologism in place of ąžuolai, by analogy with daigai ~ dāgs. Dāigs would have arisen by reanalysis from daigai: dāgs.

In some dialects, the law á: ai or á: au has also operated; cf. Naujokátis, but Jókūbaitis, Jūrgaitis;

jāns, but jaunà

oak: x = { dāgs: seedlings / šiads: straw; x = { 1. oaks / 2. aužuolai

There is also the form āržuolas (Zasėčiai, Slānimo district; Marijampolė, Garliava), which is difficult to derive from ąžuolas.

22. Lithuanian vilkai cannot be derived from Indo-European vľgai (: ϑeol, tol), with an acute ending, since such forms as būtai, pautai. contradict this. From Indo-European bhūtoí Lithuanian would have *butai; cf. tù sakai [: tu vedi ~ vedie-si < *védhei-(s)] < *sākā + éi, more likely sākā+ē > *sākāi > *sakāi > sakai.

Or must we simply return to the old explanation that vilkai is Indo-European neuter *vľgai (cf. Trautmann, Altpr. Spr. 218 sq.)?

23. In the 1605 Catechism, prieg kriksz tie 31₂₁ is apparently the loc. sg. of i-stem nouns; cf. Wieszpatie-p in Chilinskis Bible, Psalm XL, 4.

Daukša’s Wiesznati-p’ (Post. orig. 537₂₉), if it has a short t after it and a short i, may be the loc. sg. of a consonant stem; cf. gen. pl. Wieszpat-u, nom. pl. Wieszpat-es.

Vieš-pat-i-p’: Vieš-pat-ie-p’ = that is, dañti: dañtie.

24. Terpù [= terpù móteru 1605 m. Kat. 56₃, terpì saui 79₁₃, but terpu-sawi 46₁₇ terpu sawi 51₈, terpu mùsu 42₉; 52₁₀,₁₈ terpu Dētaru 85₁₈ terpu móteru 56₁₄ terpu kitu 69₁₈ < terpù with very weak proclitic stress] derives from terpu < *térpu, owing to its proclitic position; cf. Greek ὀντὶ παντὸς ἔχειν < Sanskrit ánti, περί τάτρης < περί = Sanskrit pári.

Περὶ ἔρδος II. 6₃₀₁: ἔρδος περί II. 16₄₇₆ = stands under the fence: I returned from Kaunas at night ‘having set off in the evening, I returned by morning’ Kvedarna = deléig to ne sielókites D Post., or. 330₆ mìli ne deléi W. Dievo l. c. 331₁: For God’s sake, for God’s sake I returned home. Kupiškis ‘barely, with difficulty’ (the first word [= Dievo] with very weak stress).

Besides tarp: terp ~ terpu, the following forms are also known: tarpa ~ tarpo, tarpu, tarpe, terpei, tarpe; see Bezzenb. BgLS. 56, 71, 72, 331.

Terpei Pagonų (Bretkūnas) may be the loc. sg. fem. from nom. sg. térpė (Dusetos). Tarpe is the loc. sg. of tárpas; cf. mūsų tárpę ‘between us’.

Bretkūnas’s tarpa ~ tarpo may correspond to Latvian starpā, that is, the loc. sg. of nom. *tarpa fem.

25. Daukša’s forms ik’ tam’ mėti in Post., or. 392₂₁ correspond exactly to ik tam-met in the 1605 Catechism, 44₂₅; cf. ikszammet 72₂₂.

83

Compare Daukša’s prieg dvāri (Post., or. 434₁₇) with prieg stulpų in the 1605 edition, 85₂₂ (: prieg krikštie 31₂₁).

(ik’) tam’ mėti are both locative singular forms; tam’ is from *tami, cf. tami-p berneļip in DPost., ed. Volt. 46₁₉: Avestan aē-ta-hmi, Slavic to-mb, Brugm. Gr. II², 2, p. 362 sq. tāmi-p(i) [this accent position points to Zasetiškiai kámṕ ‘at whom’, támṕ ‘at that one’. From *kamipi we would expect *kaṃp]: berneļi-p(i) [it still seems to me that in the loc. sg. i(u) was long]

= tamę: berneļį-e.

This one = Prussian *schi-s-men (schisman: a, apparently, is aᵉ or eᵃ [Baranauskis]) from Aist. this-mi + én ~ thi-smi + én. From *šīmī + én šīmjen should have arisen (cf. mes vėmja-me) > šī-menī [j after a consonant + e disappears: Aist. *svētjas ‘guest’ ~ voc. sg. *svetjē > svetė] > *šimen > šimę.

We can compare Proto-Lithuanian loc. sg. ši-m[j]-en: ši-mi with Oscan hūrtin from *hortej-en ‘in horto’: comenei ‘in comitio’, Brugm. Gr. II², 2, p. 180.

From loc. sg. ik’ gāli (DPost. or. 393₉), prieg dvāri (l. c. 434₁₇), we can very easily derive a locative with the postposition -ėn.

Gāli + én, dvāri + én > gālien, dvārięn > gālēn, dvārēn > galēn, dvarēn > galę, dvarę.

Gāli dvāri are unclear in their endings; *galē *dvārię would be expected.

26. From the Žyteliai dialect Domeikai-p ‘at Domeikis’ makes it clear that the locative singular ending -āi had a circumflex.

Me’rgāi dat. sg.: meirgā’i loc. sg.

Mergā’i + én > mergāj-én > mergoj-ę

{Latvian malā or malā does not agree with *mergājēn. Did mergā’i + én> change into *mer-gā-jén?: -āi, but -āi + én = ājen = Latvian malā ā-jen?

[11]

1910. X. 14.

The letter without its ending (up to “…valau Suv. Naumiestis”) was published in K. Būga’s “Collected Writings” (V., 1961, vol. 3, pp. 832–834) by Z. Zinkevičius, with the note: “The end of the letter is missing.” Recently, while examining K. Būga’s archival materials held in the manuscript repository of the Institute of the Lithuanian Language and Literature of the Academy of Sciences of the Lithuanian SSR (which also holds the printed letter, shelfmark F 7—30), a sheet of the same format, but of a different paper color, written in K. Būga’s hand, was found. Comparing the text of the whole letter with that of the discovered sheet showed that it was the ending of the aforementioned letter. Since the published text is readily available, only the part of the letter beginning where the etymology of the Prussian word tlaku starts is repeated here, and the discovered portion of the letter is published for the first time.

Prussian tlaku ‘he threshes’ is not a loanword from Polish tłoczyć, but a Baltic word. It is related to East Lithuanian luōkiju, luōkyti ‘to use that kind of fishing method in which boys, who break through the ice with nails and sticks, are employed,’ Dusetos, Linkmenės.

Pr. tlaku from Aist. *tlōkā (cf. lāiku per-bānda): luōkija = (I drive the horses from the field) valau Suv. Naumiestis: vāliju Kv. Riet. = žārdau Dus. Tilžė: žārdiju Kv. Riet.

84

Tlaku: luõkija = Tlokun-pelk ‘marsh on the old Prussian-Masurian border, 1347 J’: lokys lõkio ‘bear’.

The Aist. tlõk- is cognate with Slavic tolk- (Pol. tłoczyć, Russ. толочь).

Tlõk: tolk- = *ρο-τός *ρώ-γ-λη; τόρ-μο-ς ‘any hole into which a peg is inserted’ Lat. ter-m(a)-s ‘worm’ = OE wlōh f. ‘fiber, fringe, flake’ from *vlo-kā or *vlā-kā: S. Indic valkā- ‘bast’ Russ. волокно (волоть ‘fiber’).

What do you think of this etymology?

I have a lot of work. I am preparing for the final spring examinations. How are your affairs? Have they still not confirmed you as a professor?

I wish you all the best. Yours, Kaz. Būga