1909.] NATURAL SCIENCES OF PHILADELPHIA. 305 



scribed any developmental stage of the cribellum, for what Jaworowski 

 described as a "rudimentary cribellum" in Trochosa was really a 

 colulus. Fig. 33 represents a ventral view of the abdomen shortly 

 after reversion, and the abdominal segments are well pronounced. 

 The tracheal stigma (T. St.) is at the boundary of the third and fourth 

 segments. On the fourth segment are found the anterior spinnerets 

 (A. Sp.), and between them a pair of large thickenings (Or.), the paired 

 primordium of the cribellum. On the fifth segment are seen the 

 posterior spinnerets (P. Sp.), and between them the median spinnerets 

 as a pair of thickenings (M. Sp.). The resemblance of the cribellum 

 to the median spinnerets at this stage is very close, each consisting of 

 a pair of elongate ectoblastic thickenings near the mid-line; and the 

 cribellum is even larger than the median spinnerets. There can be 

 no doubt in this case of the homodynamy of cribellum and median 

 spinnerets. 



Fig. 36 shows a ventral view of the cribellum of a mature female of 

 Hyptiotes cavatus, and fig. 35 a similar view of that of Dietyna vohipis. 

 In both of these there is a distinct pair of muscles (M.) passing from 

 the spinning plate (Cr. PI.) to the anterior edge of the cuticular frame. 

 In these species the spinning plate is not paired, nor yet is the deeper 

 portion composed of the ductules paired, in which points these species 

 appear much less primitive than Filistata. I find the cribellum of 

 Uloborus plumipes to be very similar to that of Hyptiotes. In Hypti- 

 otes the tracheal stigma (T. St.) is immediately contiguous to the frame 

 of the cribellum, while in Dietyna (fig. 35) they are separated. 



Thus we come to the conclusion that the spinnerets, colulus and 

 cribellum all develop from the fourth and fifth abdominal segments; 

 the anterior spinnerets and colulus or cribellum from the fourth, the 

 median and posterior spinnerets from the fifth, as concluded by Jawor- 

 owski. The colulus and cribeilum arise as elongate thickenings mesial 

 from the appendages of these segments, the appendages becoming the 

 anterior and posterior spinnerets. Colulus and cribellum, as we have 

 seen, are in their development homodynamous with the median spin- 

 nerets of the segment behind them. Accordingly, colulus and cribel- 

 lum arising in a similar position in corresponding segments must be 

 considered essentially homologous organs, conformably with the view 

 of Blackwall, whom Dahl (1901, 1904) has followed. The case of 

 Filistata renders it a strong probability that the colulus and cribellum 

 correspond with the antero-median spinnerets of Liphistius. The 

 cribellum is a less degenerate structure than the colulus, because it 

 still possesses spinning glands, is larger, and is more frequently paired. 

 20 



