The text is full of magic experiments: experiments of theft; experiments in invisibility; love experiments; experiments in gaining favor; experiments in hate and destruction; “extraordinary experiments”; “playful experiments”; and so on. These with conjurations, characters, and suffumigations make up the bulk of the first book. The second book deals chiefly with “how exorcists and their allies and disciples should conduct themselves,” and with the varied paraphernalia required by magicians: fasts, baths, vestments, the knife or sword, the magic circle, fumigations, water and hyssop; light and fire, pen and ink, blood, parchment, stylus, wax, needle, membrane, characters, sacrifices, and astrological images. Two of its twenty-two chapters deal with “the places where by rights experiments should be performed” and with “all the precepts of the arts or experiments.” In another seventeenth century manuscript are _Seven Books of Magical Experiments of Hermes Trismegistus_. “And they are magic secrets of the kings of Egypt,” drawn, we are told, from the treasury of Rudolph II, Holy Roman emperor from 1576 to 1612.[697] Another manuscript at Vienna contains a German translation of the same work.[698] FOOTNOTES: [653] For detailed references for this and the preceding statements see Lippmann (1919), pp. 357-9. [654] I have used the edition of Paris, 1564, _Liber de compositione alchemiae quem edidit Morienus Romanus Calid Regi Aegyptiorum Quem Robertus Castrensis de Arabico in Latinum transtulit_. A number of MSS of the work will be found listed in the index of Black’s Catalogue of the Ashmolean MSS, and elsewhere, as in Sloane 3697 and Digby 162, 13th century, fols. 21v and 23r. Other editions are Basel, 1559; Basel, 1593, in _Artis Auriferae quam Chemiam vocant_, II, 1-54; and Geneva, 1702, in J. J. Manget, _Bibliotheca Chemica Curiosa_, I, 509-19. [655] See above, chapter 38, p. 83. [656] Berthelot (1893) I, 234, took the date to be 1182 A. D. and so, on the basis of this remark, placed the introduction of Arabic alchemy into Latin learning 38 years too late. It is rather amusing that Lippmann, who elsewhere avails himself of petty pretexts to belittle the work of Berthelot, should have overlooked this error. He still (1919), pp. 358 and 482, states the date as 1182 A. D., although he is puzzled how to reconcile it with that of 1143 A. D. for Robertus Castrensis or Robert de Retines. He also is at a loss as to the identity of this Robert or the meaning of “Castrensis,” and has no knowledge of the publications of Karpinski (1915) and Haskins, EHR (1915). [657] Lippmann (1919), p. 358.