Page:Oregon Historical Quarterly vol. 4.djvu/358

348 aided by the government, the business nevertheless proved unprofitable, and after selling out at a heavy loss, he came to San Francisco in 1873. Throughout life, Savage was a constant reader, with a special fondness for history. He once said that he believed he had read the histories of all the world.

From a perusal of what Nemos says concerning the History of Mexico, we are led to infer that Bancroft again wrote the introduction, as the former librarian credits his chief with two chapters of the first volume. Nemos wrote the remainder, but Bancroft rewrote some of his work, he said only a fifth, much of the revision consisting in a mere change of words. Oak differed with him on this point, holding that Bancroft did more rewriting, but Nemos persists that this is an exaggeration.

The second volume was done by Nemos, Savage, and Peatfield, Nemos writing the first half and some later chapters, two thirds of the volume in all, Savage one fourth, and Peatfield a little.

Of the third volume, Nemos wrote between a third and a half, including, as he tells us, the leading institutional and political parts, Savage a third, a writer named Griffin (George Butler Griffin was a native of New York state, and a graduate of Yale. He was a linguist, and had been an engineer in South America. Apparently early in the eighties, his connection with the library had ceased. He died by his own hand.) two or three chapters, and Peatfield a part.

Of volume four, Bancroft did one chapter, Peatfield a fourth of the whole, and Savage a third. Nemos "assisted on parts," his work aggregating a fourth of the volume.

The fifth volume of the Mexican History, embracing the period from 1804 to 1861, was known as Savage's volume. Of the manuscript, he actually wrote about