refraction are discussed for the first and only time in any extant work of antiquity,[504] provided the _Optics_ has really come down in its present form from the time of Ptolemy. His authorship has been questioned because the subject of refraction is not mentioned in the _Almagest_, although even astronomical refraction is discussed in the _Optics_.[505] De Morgan also objects that the author of the _Optics_ is inferior to Ptolemy in knowledge of geometry.[506] Possibly a work by Ptolemy has received medieval additions, either Arabic or Latin, in the version now extant; maybe the entire fifth book is such a supplement. That works which were not Ptolemy’s might be attributed to him in the middle ages is seen from the case of Hero’s _Catoptrica_, the Latin translation of which from the Greek is entitled in the manuscripts _Ptolemaei de speculis_.[507] [Sidenote: Medieval translations of _Almagest_.] If there is, as in other parallel cases, the possibility that the medieval period passed off recent discoveries of its own under the authoritative name of Ptolemy, there also is the certainty that it made Ptolemy’s genuine works very much its own. This may be illustrated by the case of the _Almagest_. On the verge of the medieval period the work was commented upon by Pappus and Theon at Alexandria in the fourth, and by Proclus in the fifth century. The Latin translation by Boethius is not extant, but the book was in great repute among the Arabs, was translated at Bagdad early in the ninth century and revised later in the same century by Tabit ben Corra. During the twelfth century it was translated into Latin both from the Greek and the Arabic. The translation most familiar in the middle ages was that completed at Toledo in 1175 by the famous translator, Gerard of Cremona. There has recently been discovered, however, by Professors Haskins and Lockwood[508] a Sicilian translation made direct from the Greek text some ten or twelve years before Gerard’s translation. There are two manuscripts of this Sicilian translation in the Vatican and one at Florence, showing that it had at least some Italian currency. Gerard’s reputation and his many other astronomical and astrological translations probably account for the greater prevalence of his version, or possibly the theological opposition to natural science of which the anonymous Sicilian translator speaks in his preface had some effect in preventing the spread of his version. [Sidenote: The _Tetrabiblos_ or _Quadripartitum_.]