AN AMPHIPOD PROM THE TRANSVAAL. 949 



two others on the opposite side of the valley. In both we found 

 a plentiful supply of watei\ 



As to these two caves, one was considerably deeper than the 

 other. In the first one we entered, we reached the level floor 

 fairly soon ; here we found shallow sheets of water supporting a 

 lai^ge number of Gammarids, Copepods, and Ostracods ; the bottom 

 was muddy and the floor of the cave generally covered by the 

 droppings of bats. No light entered as far in as this. The 

 bottom of the second cave we reached after a rather long, steep, 

 and winding descent : the stalactites and stalagmites showed it 

 to be a limestone cave. The water here was as clear as crystal, 

 rich in Gammarids and poor in Copepods ; there were no 

 Ostracods, but a few aquatic worms were taken ; some terrestrial 

 Isopods and some spiders were also found. Though the water 

 appeared quite still, a fresh supply was without doubt being added 

 continually. The bottom was rocky, covered by a thin layer of 

 fine mud, and the floor of the cave, as in the other, was strewn 

 with bat droppings. 



Later in the day a few Gammai-ids, similar in every respect 

 to the others taken, were caught under stones at a spring in 

 the vicinity, but none were ever taken in the spruit which runs 

 through the valley, where the numerous crabs would make short 

 work of such fry. 



All the Gammarids so far found in the Transvaal represent a 

 single species, belonging most probably to the genus Eucrangonyx 

 Stebbing, and closely related to Eucrangonyx vejdovskyi Stebbing, 

 vide (3. p. 389 and 4). However, this little cave form has not been 

 placed in this genus without notice being taken of certain resem- 

 blances and affinities to the genus JSHjyhargus Schiodte (3. p. 405), 

 more so to jSFeoniphargus Stebbing (3. p. 404 and 2, p. 73), and 

 the genera Crangonyx Bate and Paracrangonyx Stebbing (3. 

 p. 369 and 1. p. 218). In many ways this species appears to be a 

 generalized Niphargus-Crangonyx type, as is seen mainly in the 

 nature of the telson and third uropods ; the breadth of the second 

 joint of pereiopods 3 to 5, and the total length of these appendages 

 as compared to pereiopods 1 and 2 ; the shape and size of the 

 gnathopods ; the number of setee and spines on the inner and 

 outer plates of the maxillula ; the structure of the lips ; the size 

 of the accessory flagellum and total length of the antennule. 



It is in the sense of the comparative generalization of its 

 structure that this creature may be called primitive ; primitive as 

 opjDosed to the more recent genera Paracrangonyx, A^jocrangonyx^ 

 and Crangonyx, but, on the other hand, more recent than the 

 genus Gainmarus, and probably ^S^iphargus. This conclusion has 

 been arrived at more by the study of the telson than of any other 

 part (vide 1. p. 219). 



The large size of the outer ramus of the third uropod appears 

 to be almost peculiar, but, judging from Chilton's (1. p. 218) 

 remarks on the variability of this element in Crangonyx, this 

 character can be given apparently too much attention. 

 • Proc. Zool. See— 1911, No. LXY. 65 



