68 The Graspedosomatidse of North America. 



late as in T. album. That this is really a Trichopetalum we are 

 not confident, but judging from the exterior it is nearer to Tricho- 

 petalum than to the other genera. One of Mr. Bollman's specimens 

 is mentioned as being a male, and it may be reasonably supposed 

 that he would have i-ecognized and mentioned the conspicuous 

 and characteristic ninth pair of legs. The eyes depart somewhat 

 though not seriously, from the Trichopetalum type, for they have 

 a third row of ocelli in addition to the lunate arrangement in two 

 rows (fig. 50). 



A further difference consists in the fact that the setae are 

 smaller and arranged more nearly in a transverse line, even on 

 the anterior segments. 



Tricliopetalum (?) ocellatuin (Packard). 



Polydesrmis ocellafus Packard: Am. Nat. XYII, 428 (1883). 

 Craspedosoma packardii Stuxberg: Am. Nat. XIX, p. 400 (1885). 

 Craspedosoma ocellatum (Packard) Bollman: Bull. 46 U. S. Nat. Museum p. 

 120 (1893). 



Plate III, figures 30-35. 



" The body is small, rather short, somewhat spindle-shaped, aud unusually 

 cylindrical. There are twenty-nine segments behind the bead; the scuta^ are 

 provided witb the usual prominent lateral ridges. There is a row of dorsal 

 setae on each side of the median line; the surface of the scutes is finely granu- 

 lated. The sides (gense) of the head are full and swollen, somewhat wider 

 than tbe first scute but about as wide as the body behind. Ocelli 12-13 in 

 number, conspicuous and black. Antennte witb joint 4 half as long as 3, joint 5 

 remarkably thick, increasing in thickness to the end, longer than joints 6 and 

 7 taken together. Joint 6 not quite as long as thick; joint 7 small, conical; 

 broad at base where it is nearly as thick as the 6th; the joints rather setose; 

 there are two terminal flattened sense-sette on the tip of the seventh joint. 

 The end of the body is moderately blunt, with four sefee on each side of the 

 last segment (anal valves). It is of a pale horn color. Length 4 mm. 



' ' This aberrant form would not at first be regarded as a Polydesmus, but 

 would be easily mistaken for a Trichopetalus. The individuals were mature 

 or nearly so, as they were horn-brown. In the cylindrical body and tbick 

 antennse it approaches Polydesmus cavicola Pack., from a cave on the shores of 

 the Great Salt Lake. It differs from that species, which is eyeless, in the fusi- 

 form body, much thicker antennie, and the finer granulations as well as the 

 larger number of segments. ' ' 



The above is Dr. Packard's original description. The species 

 seems to have been placed in Polydesmus because of carinte larger 

 than those of Trichopetalum, and the granulations of the dorsal 

 surface. From the number of segments, shape of antennae, size, 



