The experimental character of John’s own handbook on astrology is worth noting. In the main, it is true, he follows the works of the philosophers and astrologers of the past, especially when he finds them in agreement.[195] Besides constantly alluding to what astrologers in general or the ancients say on the point in question, he often cites of the Greeks Ptolemy and Dorotheus (“Dorothius”) and Hermes and Doronius, but probably through Arabic mediums. He also gives us the views of the masters of India, and distinguishes as “more recent masters of this art”[196] the Arabic writers “Alchindus” and Messahala. The latter he seems to regard as an Indian or at least as skilful in their methods of judgment.[197] But he also notes when his authorities are in disagreement[198] or points out that his own experience in many nativities contradicts their views,[199] against which John’s readers are warned when they find them in the books of judgments. Even Ptolemy is twice criticized on the basis of actual experiment.[200] We see that John was not merely a translator or writer on astrology but an expert practitioner of the art. He supplements the divergent views of past authorities, or qualifies their consensus of opinion, by his own apparently rich experience as a practicing or experimental astrologer. Indeed, for him the theory and practice of the art, the paths of reason and experience, are so united that he not merely speaks of “this reasoning” or view as being “tested by experience,”[201] but seems to employ the words _ratio_ and _experimentum_ somewhat indiscriminately for astrological tenet or technique.[202] [Sidenote: Gundissalinus _De divisione philosophiae_.] The chief known work of Gundissalinus, the archdeacon who was for a time perhaps associated with John of Spain in the labor of translation, is his _De divisione philosophiae_,[203] a treatise which owes much to the Turkoman Al-Farabi (Muh. b. Muh. b. Tarchân b. Uzlag, Abû Nasr, el-Fârâbî). If Baur is right in thinking that Gundissalinus made use of translations by Gerard of Cremona, 1114-1187, in the _De divisione philosophiae_,[204] it would appear to be a later work than his translating for Archbishop Raymond, 1130-1150, which perhaps began as early as 1133.[205] [Sidenote: Place of magic in the classification of the sciences.]