honesty. The "raw material" of those numbers, with the historical references exactly given, exists in Madison's papers in his own handwriting, and is printed in his Writings, Vol. I, 293-314. Take No. 20, for example, as a test case. Fully nine-tenths of it is drawn from Madison's own abstract of Sir William Temple's Observations upon the United Provinces and of Felice's Code de L'humanité, This can be verified by any one in a few minutes by comparing No. 20 with pp. 302-309 of Madison's Writings, Vol. I. That Madison should assert No. 20 as his own was natural and right; that when Hamilton's assertion of joint authorship was made public he should explain the discrepancy by stating the facts was also natural; that his explanation was truthful internal evidence proves beyond a doubt, and that he *' conceded" No. 20 to be a joint work in any common acceptation of the term is without foundation. Sir William Temple's claim to be recognized as joint author of No. 20 is far stronger than Hamilton's. There are two paragraphs out of twenty-four in No. 20 which appear to have come from Hamilton. Most of the rest is from Sir William Temple. The case with Nos. 18 and 19 is similar, although neither is drawn from so few sources as No. 20; in each there is a possibility of a larger use of Hamilton's notes. After a comparison of these numbers with Madison's Notes on Confederacies, no editor can have any excuse for assigning these numbers to "Hamilton and Madison," as has been uniformly done by Hamiltonian editors since 1810. It should at least read, "Madison and Hamilton," although there seems to be no good reason why the exact and truthful course of the Gideon editions should not be followed in the future.
It will hardly be denied that eight of the twelve "errors" of the Madison lists now disappear, and we have then four errors in regard to two numbers in the Madison lists as compared with Hamilton's two errors in regard to two numbers.
When Mr. Lodge believed Hamilton's testimony six times as good as Madison's, he regarded the question of the authorship of Nos. 49-58 as almost evenly balanced between the
