stood that these were merely points of export, not the localities where the garnets were found. Coraes86 suggested that Μασσυλία?, a district in the country behind Carthage, should be the correct reading here. Though it is true that Strabo mentions only localities in North Africa as western sources of these precious stones, the more comprehensive list of Pliny87 includes Massilia (Marseilles) and Olisipo (Lisbon) as well. Theophrastus mentions Massalia again in section 34 as a place from which precious stones were obtained. On the whole, the conjecture of Coraes is not plausible enough to justify an alteration of the text.
19. The stone found near Miletus does not burn; it is angular and there are hexagonal shapes on it. It is also called anthrax, and this is remarkable, for in a way the nature of adamas is similar.
The kind of anthrax found near Miletus on the western coast of Asia Minor evidently had a striking and peculiar form. Theophrastus appears to be speaking of a well-crystallized mineral with hexagonal facets. It is significant that both garnet and spinel often occur crystallized in this manner. Of the two, spinel seems more likely, since the anthrax mentioned in the present section is apparently different from the kind described in section 18, which was almost certainly garnet. Since the anthrax of this passage is said to be like adamas, it is even more likely that it was spinel. Though adamas seems to have been a general term used for several minerals that were unusually hard, the descriptions of Pliny88 suggest that it generally referred to corundum, particularly the mixture known as emery; this frequently contains spinel in addition to corundum, and almost invariably magnetite, which is very similar to spinel in crystal form. The occurrence of large deposits of emery in Asia Minor not far from the site of ancient Miletus supports the relationship between this kind of anthrax and adamas. Most of the emery in these deposits contains only about 50 per cent of corundum, and the remainder consists of the associated minerals.89 Though it is remarkable that this particu-
86 Ed. of Strabo, IV, 357. 87 XXXVII, 92-97.
88XXXVII, 55-61; see also the notes on sec. 44.
89 C Schmeiszer, Zeitschrift fur prahtische Geologic, XIV (1906), 188.