but we cannot afford to forget that neither Vararuchi nor Grimm nor Verner, nor all of them taken together, can be wholly relied upon to explain all the deviations from the norm. How the ear of a man will be the recipient of a sound, or how he will imitate it in speech, will depend upon his culture; what the "apabhramśa" will be in one stage of culture, will not be so in another. Consequently, the generalised rules of equation applicable to some words of one speech, may not be applicable to other words of that very speech. There are also other good reasons why we cannot acknowledge the all-sufficiency of the rules alluded to, but it will be a digression to adduce them here. What I want to bring out prominently is that, we cannot study the phonetic changes in a speech without taking the speakers of it into account. After observing the differences among the sister dialects, we raise the question, why the parent tongue underwent different sorts of changes in different provinces, we ask why the "apabhramśa" forms in use in Hindi for example, did not become current in Bengali? What were the solvent elements in different provinces that brought about the characteristic changes noticeable in different speeches of common origin? To get to the facts, which induced different sorts of changes and modifications in different provinces, we must direct our attention to provincial racial peculiarities, as well as to the physical conditions of life, which were present in those provinces. This is exactly what is not done by many philologists. And we shall presently see how they create imaginary races to explain away their difficulties without caring to study the actual racial peculiarities existing in different provinces. This is the reason why many scientists look to the philologists and their work with much disfavour. Such an eminent man of science of our time as Karl Pearson spaeks slightingly of the philologists, as they do not generally
