Historical, nay, even individual, influences come in which prevent the science of language from becoming purely mechanical. Pott, and Curtius, and others stood up against Bopp and Grimm, maintaining that there could be nothing irregular in language, particularly in phonetic changes. If this means no more than that under the same circumstances the same changes will always take place, it would be of course a mere truism. The question is only whether we can ever know all the circumstances, and whether there are not some of these circumstances which cause what we are apt to call irregularities. When Bopp said that Sanskrit _d_ corresponds to a Greek δ, but often also to a Greek θ, I doubt whether this is often the case. All I say is, if _deva_ corresponds to θεὁς, we must try to find the reason or the circumstances which caused so unusual a correspondence. If no more is meant than that there must be a reason for all that seems irregular, no one would gainsay that, neither Bopp nor Grimm, and no one ever doubted that as a principle. But to establish these reasons is the very difficulty with which the Science of Language has to deal. There is no word that has not an etymology, only if we consider the distance of time that separates us from the historical facts we are trying to account for, we should sometimes be satisfied with probabilities and not always stipulate for absolute certainty. Many of Bopp’s, Grimm’s, and Pott’s etymologies have had to be surrendered, and yet our suzerainty over that distant country which they conquered, over the Aryan home, remains. If there is an etymology containing something irregular, and for which no reason has as yet been found, we must wait till some better etymology can be suggested, or a reason be found for that apparent irregularity. If the etymological meaning of _duhitar_, daughter, as milkmaid, is doubted, let us have a better explanation, not a worse; but the general picture of the early family among the Aryans “somewhere in Asia” is not thereby destroyed. The father, Sk. _pitar_, remains the protector or nourisher, though the _i_ for _a_ in _pater_ and πατἡρ is irregular. The mother, _mâtar_, remains the bearer of children, though _mâ_ is no longer used in that sense in any of the Aryan languages. _Pati_ is the lord, the strong one—therefore the husband; _vadhû_, the yoke-fellow, or the wife as brought home, possibly as carried off by force. _Vis_ or _vesa_ is the home, οἰκος or _vicus_, what was entered for shelter. _Svasura_, ἑκυρὁς, _Socer_, the father-in-law, is the old man of the _svas_, the _famuli_, or the family, or the clients, though the first _s_ is irregular, and can be defended only on the ground of mistaken analogy. _Bhrâtar_, _frater_, brother, was the supporter; _svastar_, _soror_, sister, the comforter, &c.