operation of these rules be at once transformed into Sanskrit as Deśaja. European scholars also have got no further than Prakrit, and seem to believe that the modern dialects are merely corruptions of Prakrit forms. It is therefore not presumptuous to say that further research, and a better acquaintance with the laws of development of these languages, will probably enable us ere long very much to diminish the number of these Desajas by tracing them back through newly discovered processes either to Sanskrit or Prakrit. Even as matters stand at present, if all the Tatsamas and all the Tadbhavas be Aryan, there is only a very small proportion left which can be non-Aryan.
§ 5. Of the three classes into which all the languages have been divided in the preceding section, Tatsamas are the least interesting to the student. This class consists of pure Sanskrit words which had long been dead and buried, so to speak, when in comparatively recent times they were resuscitated and brought into use by learned men, partly to supply real wants, but still more to show off their own learning. They have not been current in the mouths of the people long enough since their new birth to have undergone any of those processes of change to which all really living words in every language are constantly subjected; and a great many of them, especially in Bengali and Oriya, are not likely ever to be used colloquially. They ought certainly to be excluded from dictionaries.
It is to the Tadbhavas that we must turn if we would become acquainted with the secrets of the phonetic machinery of the Aryan Indians. Of these there are two sorts, so distinguished from one another that it is impossible to mistake them. The one class consists of those words which were in use in Prakrit, and in which the Prakrit processes have been carried one step further. The other contains words which apparently have not come through Prakrit, as they exhibit a more perfect form, and
