[Of the four laryngeals assumed by the author for Indo-Hittite, three (two glottal stops and x) were clearly voiceless. This paper assembles the evidence that the fourth laryngeal (γ) was voiced.]

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.

[The formative -ka of Gk. ἕστηκα ‘I stand’ and Toch. tākā ‘I was’ must contain the Indo-Hittite 1st sg. perfect ending -xa. The fact that early Greek κ-perfects are nearly all from ‘heavy bases’ shows that in these words -xa was preceded by a laryngeal; it turns out to have been either ' or ̣̕. The behavior of the Indo-European noun suffix -k(o)- indicates that it represents IH -̣̕x-, where the -̣̕- is the collective suffix and -x- the suffix that in Hittite yields factitive verbs and verbal nouns.]

[Greek perfects in -ϕα or -χα from verb-stems ending in π, κ, or γ owe the aspiration to Indo-Hittite perfect endings beginning with a voiceless laryngeal: 1st sg. -xa and 3d pl. .]

Cet ouvrage fut subventionné en l'année 1935 par le Council of Research in the Humanities of Columbia University; il fut commencé, poursuivi et achevé sous la direction immédiate de M. L. H. Gray, professeur de linguistique comparée à la même université.

[An inspection of the known Gaulish vocabulary makes it possible to tabulate the Gaulish correspondences of IE sounds in most positions without recourse to hypothetical Gaulish forms.]

For a full century the Latin language has been subjected to the most searching scrutiny by comparative methods, semantic data being employed as a sort of control. While these studies may be judged to have reached their limits, a residue of words remains for which no satisfactory etymology has been found, or none at all, as may be discovered by consulting the very conservative dictionary of Ernout and Meillet. Upon these more obscure words, the writer believes, some light may be thrown by following semantic clues, and the following notes are offered in the hope of demonstrating this to be true. Documentation is not presented for usages that may be verified in any lexicon.

[Merely determinative compounds, though common in many modern IE languages, represent a rare type in Primitive IE. Of the few Greek examples found as early as Homer, adjectives with παv- are the most characteristic. Their secondary origin from regular ‘exocentric’ types can still be traced in the epic. This is typical for the development of determinative composition in general, in Greek as well as in other IE languages.]

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.

[The Slavic languages, with the exception of Bulgarian, in contrast with the languages of western Europe, have preserved the IE nominal inflection but have simplified the verbal system. Russian has gone farthest in this direction. Original auxiliaries in Russian have become particles; all finite verb forms except those of the present tense have been replaced by nominal or participial constructions.]

[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 paper formulates a technique for the analysis of morphological and syntactic phenomena, by means of which the Russian noun categories of declension, class, and gender are established and their mutual relationships described.]

[Certain syntactical similarities between Celtic and Ancient Egyptian—specifically, phrases of anticipatory emphasis, suffixed pronouns, and adverbial predicates—are interpreted as evidence of a general Mediterranean substratum in Celtic.]

[Only on the basis of a large collection of actually occurring examples is it possible to set up a precise and adequate definition of mahogany (or any similar word), as it is used in present-day colloquial American English. The definition so arrived at is contrasted with a typical dictionary definition based on secondary sources.]

[The long forms (ahtoie, haloian, uuacoiande, etc.) of verbs of the second weak conjugation in Old Saxon are not of IE origin as is generally supposed, but owe their existence to an attempt to make a sharper formal distinction between indicative and subjunctive by adding the subjunctive endings of the first or jan-class to the stem of the second class. The extension of the long forms to the infinitive and present participle was due in part to the metrical advantages of these forms in certain types of alliterative verse.]