Page:A Comparative Grammar of the Modern Aryan Languages of India Vol 1.djvu/362

340 languages, the stronger of the two elements holds its place, or rather has a tendency to do so, and the weaker inclines to be assimilated, and subsequently disappears.

In this nexus, moreover, there are two further sources of irregularity; the first, that the semivowels y and v, as usual, oscillate backwards and forwards, appearing at one time hardened into j and b, at another relaxed into i and u; the second, that this form of nexus is more liable than any other to be dissolved by the insertion of a vowel between its component parts.

We shall now go through the various descriptions of weak nexus seriatim in the order given above.

§ 87. —This combination is from the nature of the case very rare. It is of course impossible in Sanskrit for two nasals of different organs to be joined: ञ and ण could not combine. Therefore the only combinations which, as far as I am aware, actually occur are न + म and म + न. With regard to the former of these, Vararuchi directs the assimilation of the preceding to the following letter: जन्म becomes जम्मो, and मन्मथ by a curious set of changes वम्महो (iii. 43). The latter word does not occur in modern times save as a Tatsama. जन्म is always dissolved into जनम्, and is in this form a very much used word. P., however, has in addition the curious variations जंम, which is the Prakrit form above, and जंमण, as also the vulgarism जरम in the spoken language. Such formations as might be expected to occur from the addition of the formative syllable मन् to roots ending in न् do not survive to modern times, because the modern languages in the vast majority of Tadbhavas retain nothing of the Sanskrit elements beyond the root, and form their secondary words by additions of a different type. Roots like तन्, मन्, हन्, occur often as the basis of modern words, but with entirely modem afformatives. Similarly in the case of the nexus म्न I have not found modem instances. निम्न