Messahala[53], of which copies are, I have no doubt, sufficiently numerous. The Cambridge Library has four, viz. Hh. 6. 8, Ii. 1. 13, Ii. 3. 3[54], and Kk. 1. 1, and there is another copy in St. John's College Library, Cambridge, marked F. 25. The title should be particularly observed; for the treatise is distinctly divisible into two separate parts, viz. the 'Compositio Astrolabii' and the 'Operatio Astrolabii.' The former begins with the words--'Scito quod astrolabium sit nomen Graecum,' and explains how to make an astrolabe, and how to inscribe on it the various necessary lines and circles with sufficient exactness. It is much the longer portion of the treatise, and (in MS. Ii. 3. 3) is illustrated by numerous diagrams, whilst the second part has no such illustrations. But it does not appear that Chaucer made any use of this former part, as his astrolabe had been procured ready-made. The second part of the treatise, or 'Operatio Astrolabii,' begins with the words 'Nomina instrumentorum sunt hec.' This is evidently one of the sources from which Chaucer drew largely[55]. Chaucer's Part i. is almost wholly taken from this, but he has expanded it in several places, with the evident intention of making it more easy to understand. In Part ii. he has taken from it, with more or less exactness, sections 1-3, 5-8, 10, 11, 13-18, 20, 21, 24, 25, 27-31, 33-37, 41 and 42; whilst sections 4, 9, 12, 19, 22, 23, 26, 32, 38-40 and 43 do not appear in it. In other words, Messahala's treatise accounts for thirty-one conclusions out of forty-three, or about _two-thirds_ of the whole. In some places, Chaucer has translated almost word for word, so as to leave no doubt as to his authority. Besides which, I have already remarked that Chaucer's version is _directly_ connected with Messahala by the quotations from the latter which appear in MS. E.; see description of this MS. at p. lix. If it be inquired, whence did Chaucer derive the remaining third of his Second Part, I think it very likely that some of it may be found amongst the varied and voluminous contents of such a MS. as Ii 3. 3, which is a sort of general compendium of astronomical and astrological knowledge. The complete solution of this question I leave to some one with more leisure than myself, being satisfied that to have found the original of Part i. and two-thirds of Part ii. is to have made a good start. It must not be omitted, that the MSS. of Messahala are not all alike; that some copies have propositions which are not in others; and that the order of the Conclusions is not invariable. The chief noteworthy difference between Chaucer's version and the Latin original is in the order of the Conclusions; it is clear that Chaucer not only took what he liked, but rearranged his materials after his own fashion.