Page:On the Difficulty of Correct Description of Books - De Morgan (1902).djvu/15

 or the other. It happened that all the four gave us what we wanted to illustrate.

The first book was a collection of four geometrical and two astronomical treatises by John Werner of Nuremberg, quarto, 1522, beginning "In hoc opere haec continentur. Libellus Joannis Verneri Nurembergen. Super vigintiduobus elementis conicis " [7] lt is said that this book was so rare in the time of Tycho Brahé, that he could not find it in all Germany, though he secured a copy at last in Italy. The two last treatises being astronomical, we turn to Lalande's 'Bibliographie Astronomique,' and we find at the right year, 1522, that this book consists of the two astronomical treatises, followed by an epistle of Regiomontanus to Cardinal Bessarion on the meteoroscope {instead of preceded by four geometrical treatises of Werner himself]. The authority is Weidler, who, says Lalande, adds two other tracts as contained in this work, of which Scheibel observes that they have never been printed at all. Here is a heap of confusion, in which three noted writers of mathematical history are concerned. Looking at Weidler (at the page cited), we find reason to think the case stands as follows. Weidler, after hinting that Werner printed the works of others as well as his own, gives a list as extant, in which he takes no care to distinguish between what Werner only printed, and what he both wrote and printed. In the middle of this list comes the letter to the cardinal. The last five of the list are five of the treatises which really are in the work before us, the sixth being omitted. Then, says Weidler, these last five works appeared at Nuremberg in 1522. From this it would appear as if Lalande had selected two astronomical works of Werner, the letter of Regiomontanus, and two others which he does not name because Scheibel said they were never printed.

We had turned to Weidler's History, because La-