ority of discovery, of adaptation, or of introduction. A printer may thus lose his character as an artist; he may be judged in 1852 by his type of 1842; similarly, the skill and knowledge of the author in 1852 may be set down as being what they were in 1842. It often happens that the author has no knowledge of what has been done: the edition may pass from the hands of the original publisher into those of another, with whom the author has nothing to do. Sometimes the alteration is made by the original publisher. (17)
Secondly, a title sometimes undergoes alteration which, whether by intention or carelessness, gives an account very different from the truth. We have before us a book, in which the genuine title describes it as containing matter from 1700 to 1846, mostly German; the substituted title describes it as containing all matter up to 1846, in Germany and the adjacent countries.
Thirdly, even in the original title, it is not uncommon to make the date a year later than the actual date of publication. When the book is published in the last months of the year, so that the right date will soon make it appear a year old, the next date is frequently used. Authors should look to this practice, by which their priority may be seriously compromised. Fifty years hence, a discovery, or other matter of merit, under the date 1851, will certainly be held to have preceded the same under the date 1852. But if a publication made in September 1851, be dated 1852, there is time for another to republish it under the date 1851, and thus, with or without intention, to secure it in future history. From the preface of a Latin edition of Wallis's Algebra, it appears that this practice of advancing title-pages was common in the year 1685. The truth is, that the year alone is not now definite enough: every title-page should bear the month of publication, as well as the year. It would also be of much advantage, if there were an
17
