Tertiary Spiders and Opilioiiids. 221 



above, characters the combination of which might have come to 

 our help, not a single one can be recognized. Even in recent 

 spiders the spiracle can be seen only with difficulty, the diurnal 

 tvpe of eyes is an expression of the internal structure and cannot 

 Ije preserved, the clypeus, if properly preserved would appear only 

 in a plane vertical to the surface of the specimen and the absence 

 of the colulus is only a negative character. The claws could be 

 seen under microscope if they were preserved, but although every 

 spider has either two or three claws on each tarsus, none of the 

 specimens in question show any trace of them. Similarly, 

 although the specimens show well preserved hair and spines, there 

 is no trace of scales. The evidence, such as there is, points there- 

 fore against the probability of Parattus belonging to the Family 

 Attidffi. 



But let us assume for the moment that Parattus is an Attid. 

 The separation of Attidaj into genera is based on the following 

 characters : — -i ) Number and structure of teeth on the lower 

 margin of chelicerce ; 2) Size and shape of the entire eyegroup; 

 3) Relative position of the eyes of the second row; 4) Xumber of 

 spines on the underside of the first tibia; 5) Number and distri- 

 bution of other spines on legs; 6) Distance between the anterior 

 coxae; 7) Shape and especially elevation of carapace; 8) Shape 

 of underlip ; and 9) Shape of sternum. Most of these structures 

 are either poorly preserved or entirely indiscernible. If the family 

 relationship were definite then indeed the genus would be suffi- 

 ciently characterized by the proportion of the anterior e3'es and we 

 could disregard the other characters. As for the specific dififer- 

 ences given by Scudder for his three species of Parattus and con- 

 sisting in the shape and proportions of the cephalothorax, these 

 differences in recent spiders would undoubtedly have generic and 

 not specific value. For the recognition of species in Attidse is 

 based chiefly on the structure of the male palpus and female 

 epigynum, the distribution of color on the abdomen, carapace and 

 legs, the type of hair and scales, modification of the legs, etc. 



The case of Titanoeca ingcrma (as well as that of Titanoeca 

 hesterna) is still worse. The genus Titanoeca has been estab- 

 lished by Thorell for spiders which belong to the division Cri- 

 bellatse. Scudder mentions the fact that this genus "is not far 

 removed from Amaurohius," but in his description of both species 



