From the PREFACE.
IN giving up my first plan, I was influenced not a little by my belief, that from a comprehensive work such as this, a work in which it is sought to present the facts and problems of language in connected form, more might reasonably be expected than what I had at first proposed. A student might fairly ask that the many questions which await an answer should not be simply avoided, but that some honest attempt should be made to advance a step towards their answering. It must surely be useful that he should not only read that which can be called certain, not only be taught well established facts, but that he should at the same time find the various problems and puzzles, with which the study of Indo-Germanic inflexion abounds, at least briefly mentioned and conveniently arranged. So will the scholar guard best against the mistake which not the best scholar is wholly free from, the danger that in trying to bring order and light into his palace of knowledge, he may leave some dark riddle unattempted, and only move it from one corner to another. If amidst these shifting theories I have often taken a decided stand, and declared myself for one or other of them, adding therewithal other and many new views and explanations, I am yet far from believing that I have placed beyond all doubt the view which I have preferred. In these matters to indicate a path for future research or simply to establish a prima facie case is far harder than most people think; and many a theory which seemed to be fixed on the firmest foundation and to offer no point to attack, has been broken down in the end. I can only hope that the mistakes which these volumes must inevitably contain, may help to supply the means for their own correction.
A few of my readers perhaps may wonder why certain new and some very recent theories upon Ablaut, proethnic Accent, formation of Roots and Suffixes, and other such matters, have in these last volumes been either altogether disregarded or only just glanced at. A good deal of the most recent work I would indeed have included in this last volume but that it had to be finished in 1891. In other cases I saw before me hypotheses, which attractive as they are, and fruitful as they may prove to be, at the time of their publication were too slightly worked out by their authors, and had been too little tested to allow of my making them the basis of my own account. In this volume I have practically not touched the newest theories of Ablaut; I confess that I approach the glib and symmetrical systems of Ablaut Series (cp. Bartholomae in Bezzenberger's Beitr�ge, XVII 105) with very little confidence, and I must refer to what is said on this matter in Vol. I �309. Even a question of Verb Morphology so important as the form of proethnic Roots (whether they were monosyllabic or not) I have left on one side; I believe neither the one thing nor the other, but only that in the present state of our knowledge we can know nothing about it.