The Graspedosomafidae of North America. 13 



families. When a specimen of Trichopetalum is uncovered no 

 perturbation is apparent unless the animal is actually touched. 



Not only do the antennae come in contact with an object, but 

 also as a usual thing, the front of the head. The creatures are 

 not able to overcome their momentum with sufficient quickness, 

 or else their mental operations are too slow. They are not, how- 

 ever, worse off than many animals of higher organization, and 

 better developed sense-organs, for many reptiles and mammals do 

 not, as a rule, turn out for an obstacle until they have come into 

 actual contact with it. 



That the eyes of the Diplopoda are incapable of forming im- 

 ages is, however, chief! \^ to be inferred from their plan of struc- 

 ture, as Lubbock has pointed out. And with the flattened eyes 

 of lulidse and Polyzonidse this is at once plain ; but the eyes of 

 Craspedosomatidae are much more convex, regular and well-de- 

 fined than those of the others, and approach more nearly the con- 

 dition where effective vision might be possible. 



One of the most noteworthy characteristics of Trichopetalum 

 exists in the great freedom of motion enjoyed by the head. In 

 other families the head is set more or less deepl}'' into the first 

 segment which is hollowed out in front to receive it, and is con- 

 sequently incapable of more than a slight motion. In such cases 

 the head is smaller, that is, narrower than the first segment ; 

 but in the Craspedosomatidse the head is much broader than the 

 first segment, against which it is fastened as to a neck, and upon 

 which it moves with considerable liberty in any direction. 



In like manner the mandibulary stipes are in this family excep- 

 tionally developed, and capable of rapid movement in the pro- 

 cess of chewing. The movement of the stipes in chewing has not 

 been noticed in other Diplopoda, though its absence we have in 

 some cases observed. 



The antennae are, in most species of the family, exceptionally 

 slender, but do not noticeably differ from the diplopod type. In 

 Trichopetahim, however, they are shorter and somewhat clavate. 

 They are manipulated in the usual manner, though their motions 

 are more than ordinarily quick. 



If the antennae are dipped into a liquid or touched against any 

 thing that sticks to them they are at once cleaned in the mouth. 

 How the animals manage to keep the other parts of their bodies 

 clean is an interesting question, for although they live in places 



