It is for good reasons that the IE heteroclitic nouns, that is, the type Skt. yakr̥t, gen. yaknáḥ, Gk. η̇̑παρ: ἤπατος, Lat. iecur: iecinis (cf. iecinoris), arouse the special interest of linguists. Scholars have discussed the origin of this puzzling category of words, and have tried to enlarge the number of heteroclitic paradigms. The methodological aspect of these attempts, it seems to me, needs a closer examination; but here I want only to point out a case which might be added to the list of heteroclitic words.

In classical Old Irish the enclitic emphasizing pronoun of the third singular masculine and occasionally neuter as well as of the plural in all genders is som. Towards the end of the period, this is sometimes spelled sum or sam. Very rarely do sem or sium occur after palatal consonants. Though the vowel may vary, the final letter is almost always m. So far, only two exceptions have been noted and both are found in sentences written in the margin of the Würzburg Glosses by the main hand which may be approximately ascribed to the middle of the eighth century about fifty years later than the prima manus. In the one instance, Wb.9b7 reads a thindnacul sa 'his being delivered up', for which the editors of the Thesaurus Palaeohibernicus suggest the correction a tindnacul sa or a thindnacul som, the second emendation being supported by á anim som ‘his soul’ in the same gloss. In the other instance, Wb.9b17 reads a mbésa sa 'their customs', where instead of sa one again expects som.

[An attempt to show that the idea of unspecified place is not a matter-of-course concept, but an artificial abstraction unknown to many languages outside modern Europe.]

The purpose of this paper is to describe the accent and quantity systems of the Serbo-Croatian literary language as they function on the synchronic phonological plane, without considerations of morphology or of linguistic history, these being other and separate problems.

The description is based on data presented in the standard works, and on my own observations during a month in Yugoslavia in the summer of 1938.

In Language 15.105 Professor Rositzke draws attention to the question of the effect of the initial consonant upon the duration of the vowel which follows it. There is a tendency in some quarters to find that initial voiced consonants are followed by longer vowels than are initial voiceless consonants, and to seek to explain this as a 'natural physiological compensation between stop-energy and vowel duration', with the implication that the voiced consonant is weaker and hence followed by a longer vowel, by way of ‘compensation’.

The purpose of this paper is to show the probable point (or points) of contact between the strong and the weak verbal systems, which resulted in the development of a new analogical weak preterite form with dental suffix (which either existed alongside the original strong form or entirely displaced it). Cases of borrowing from an already existent weak verb (such as sϕkta ‘sank’ for sǫkk, hangða ‘hung’ for hekk) will therefore be omitted. It is, however, often doubtful whether the weak preterite was thus borrowed or was due to association between the strong and the weak verbal systems.