Page:The Library, volume 5, series 3.djvu/208

 196 PROBLEMS OF THE ENGLISH The table will show at a glance the provisional arrangement at which we have arrived. 1 Having now constructed a working hypothesis it remains to see whether it will work. Our scheme fits the facts so far as we have observed them, and I believe it alone will fit them. By the rules of the game it should account for all the variants in the different texts ; unfortunately, through the per- versity of actual conditions, it does not in the least follow that it will. Let us see. On the present occasion I shall, of course, be able to treat the question only in the most sum- mary manner. All I can do is to give statistical figures of the readings which support our scheme, and to consider briefly the most important of those readings which appear to contradict it. There are, to begin with, in every manuscript a number of readings in which it is opposed by a con- sensus of all others. The numbers of such readings in the different manuscripts are : P 97, H 38, 3 46, D 27, W 53, K 39 ; total 300. There are also 17 cases in which two or more manuscripts differ from the rest without agreeing among themselves. The total number of variants recorded being 424, there remain 107 cases in which they fall into groups, and are therefore capable either of con- firming or contradicting a scheme of relationship. 1 The table must not be taken to imply that, for instance, either B or S is immediately derived from /3, but merely that they are derived from it, and that the intermediate steps, if there were any, are now lost. Similarly in the text, if I speak of errors or emen- dations introduced by D, I mean, of course, by D or some ancestor of D subsequent to