INTRODUCTION. 87 which you refer. The postposition occurs in nearly all Indo-Aryan languages. Thus, in various languages, we have kā, kā, kaī, kahā, Kahū, Kaū, Kū, and other similar words. All these have certainly a common origin. The oldest forms are kahi and kahā, and these are survivals from Apabhramsa Prakrit, in which they also occur. Kahi and Kahū are simply the Prakrit forms of the Sanskrit Artè, for, just as the genitive postpositions–er, kar, kēr, kā, k, and so on are all derived from the Sanskrit Artah, which in Buddhist Sanskrit e. g. Mahāvastu, is regularly used to mean ‘of’, as in udayama-kortú-asanā, the seats of the garden.” Sir G. Grierson has dealt with this point elaborately in an article entitled “On certain suffixes in the modern Indo-Aryan vernaculars” published in Zeitschrift Fur Vergleichende Sprachforschung Auf Dem Gebiete. Der Indogermanischen Sprachen. But with due deference to his opinion I take the liberty to refer my readers to the suffix as used in the dative and accusative cases on pp. 36, 37, 40, 42, 43, 44, 53, 68, 71, 98, 617,618, 633, 701, 705, 709 and 710, and solve the problem for themself. This suffix I still hold to be of a pleonastic origin. Mr. Anderson writes on this point:- - “It may interest you to know that Koch and Kāchari have an objective in coil, which is curiously like the Hindi. Now Koch and Kāchari were agglutinative till quite recent times and certainly borowed their inflections from some Aryan language. The only Aryan language with which they have been in contact has been the ol. Is it not curious that they should have taken the objective in Coi o Take again Assamese which has an objective in E, oto-to the man.” The form Co in Bengali was originally of and the coil in Koch and Kāehari mentioned by Mr. Anderson is certainly derived from this source. The Mio of Assamese is to be traced to the old Bengali form xiào, and all these I am inclined to connect with the pleonastic *. As far as the dative is concerned I find from the table given in the article of Sir George Grierson, already referred to, that the apabhrainsa dialects as well as nearly all of the modern Indian vernaculars such as Old Gujarāti, Mod. Gujarātī, Mārwārī, Jaipurī, Naipālī, Danga, Braj Bhākhā, Standard Hindi, Panjābi, Käsmiri, Indus Köhistān, Lohndi Sindhi, Oriyā, Bengali (Marathi), Assamese, Behāri and Eastern Hindi, have similar postpositions; all of these should evidently be traced to a common origin. This may either be the pleonastie of as I hold, or the word go as Sir George Grierson has supposed. There is much truth in what Mr. Anderson says in this connection. “The subject is one that requires more investigation, and in investigating we ought not to neglect the languages on the Bengal border.”
