^{page 490 note 1} The system of writing Kurdish in Roman script here used is that given in the second of two papers which I published in the JBAS some years ago: ‘Suggestions for the use of Latin Character in the Writing of Kurdish’ (January, 1931), and ‘Some Developments in the Use of Latin Character for the Writing of Kurdish ‘ (July, 1933). For the present purpose it will be sufficient to explain that all the consonants have approximately their English values except that x represents c and j have their modern Turkish values of English j and French j respectively, and the digraphs Ih and rh represent a velar I and a rolled r which are carefully distinguished from ordinary I and of the vowels, a, o, o, and e (the open sound, not the diphthong which is written are always long, i is a neutral vowel, y is pure short i as well as the consonant, u is always short; long i and long u are represented respectively by iy and uw yi and wu after a vowel). Hyphens have been freely used to separate the component parts of words ordinarily written together in order to make the argument as clear as possible. Senator Taufiq Wahbi Beg, my debt to whom in connexion with my Kurdish studies at all stages I cannot acknowledge too often, has kindly read through this paper and sent me some comments by post. It has not been possible to discuss them with him personally and on two points of substance I have preferred to give his views in footnotes rather than seek to recast my original paragraphs. I have to thank him also for establishing the rule given in § 49 (my own recorded examples being insufficient to formulate it with confidence) and for some additional examples elsewhere. I am also indebted to Professor W. B. Henning for valuable criticisms and suggestions, especially in connexion with certain problems of terminology.

^{page 491 note 1} I am fully aware that the affixes given in column 2 are identical with the verb ‘to be’, but for the reasons given in § 43 below it is convenient to show them in a separate column.

^{page 496 note 1} I should have been tempted to link te with obsolete de (§ 19) on the analogy of pe with be, but Taufiq Beg does not consider that there is any connexion.

^{page 501 note 1} My list of observed examples shows that the inflexion is seldom thrown forward on to an adverb or prepositional phrase expressing time, manner or place, but one or two exceptions made me hesitate to try to formulate a rule more precise than that given in the text. Taufiq Beg states that the inflexion is not attached to an adverb or prepositional phrase if there is any possibility that the adverb or last word of the phrase might be misunderstood as the direct object of the verb.