This is, as we have confessed, a plausible argument, but it is at the same time a thoroughly unsound and superficial one. It skims the surface of linguistic phenomona, without penetrating to the causes which produce them. It might pass muster, and be allowed to determine our opinions, if the analytical tendency alone had been active since our knowledge of language began; if we had seen old forms worn out, but no new forms made; if we had seen words put side by side to furnish analytic combinations, but no elements fused together into synthetic union. But we know by actual experience how both synthetic and analytic forms are produced, and what are the influences and circumstances which favour the production of the one rather than of the other. The constructive as well as the destructive forces in language admit of illustration, and have been by us illustrated, with modern as well as with ancient examples. Both have been active together, during all the ages through which we can follow linguistic growth. There have never been forms which were not undergoing continual modification and mutilation, under the influence of the already recognized tendencies to forget the genesis of a word in its later application, and then to reduce it to a shape adapted to more convenient utterance; there was also never a time when reparation was not making for this waste in part by the fresh development of true forms out of old materials. Nor has the tendency been everywhere and in all respects downward, toward poverty of synthetic forms, throughout the historic period. If the Greek and Latin system of declension is scantier than that of the original language of the family, their system of conjugation, especially the Greek, is decidedly richer, filled up with synthetic forms of secondary growth; the modern Romanic tongues have lost something of this wealth, but they have also added something to it, and their verb, leaving out of view its compound tenses, will bear favourable comparison with that which was the common inheritance of the branches. Some of the modern dialects of India, on the other hand, having once lost, in the ordinary course of phonetic corruption, the ancient case-terminations of the Sanskrit, have replaced them by a new scheme, not
