all of his own writing on this work and British Columbia together at a third of a volume.
A review of the facts shows that if we exclude the comparatively few interpolations and changes made by Mr. Bancroft, we can with assurance declare the authorship of all portions of the third volume of Central America, of the volumes on California, and of those on the North Mexican States, Arizona and New Mexico, Nevada, Colorado, and Wyoming, the Northwest Coast, Oregon and Washington, Idaho and Montana, and that we can give in general terms, though without being able to locate the exact parts done by individuals, the names of the authors of two volumes of Central America, and all of Mexico, Utah, British Columbia, and Alaska. In these works Oak and Nemos were agreed that there were scattered fragmentary bits aggregating several volumes so worked over by different writers in different ways as to render it impossible to determine the exact authorship.
Turning to a consideration of the individual field of writing, we find that of the twenty-eight volumes of history proper, Bancroft is to be credited with four, no one entire, Oak with seven and a half, Nemos five, no one entire, Mrs. Victor a little less than five, Savage over three, Peatfield one and a half, principally in small parts, and Bates one and a fourth. (This is a computation based exactly upon the facts as given, except in Bancroft's case.) Nemos upon the same basis makes the shares, except Savage's and Bancroft's, all slightly greater. He assigns to Oak between seven and a half and seven and two thirds volumes, to himself and Mrs. Victor over five each, to Peatfield about two, and to Bates one and a half. An actual count of the parts of volumes written by Bancroft gives a total of three and a half, but Nemos said that he took four as the number upon the authority of Oak. This would allow him a half
