144 THE VOYAGE OF H.M.S. CHALLENGER. 



lacra does not extend beyond the ambitus. These deflections give the test an angular 

 outHne when seen from above (PI. XXVII. figs. 1, 2; PL XXXV.^ figs. 9, 10), the 

 median interambulacral spaces bulging beyond the general outline. The sides of the 

 test slope similarly to the anterior extremity from the apex to the ambitus ; the median 

 ridge connecting the posterior planes is quite rounded and passes into an arched surface 

 above the anal system, while the junction of the lateral planes of the test, and the 

 flattened anterior extremity form quite well-marked slightly-rounded angles corresponding 

 to the outer edges of the anterior extremity of the test (PI. XXVII. figs. 1, 3; PI. 

 XXXV. ^ figs. 9, 11). The general trend of the outline of the test as seen from above 

 tapers very gradually towards the anal extremity from the anterior edge of the test 

 (PI. XXVII. figs. 1,2; PI. XXXV. figs. 9, 10) to the line of the lateral posterior interam- 

 bulacra ; here it bulges out somewhat, and from the median line it slopes more rapidly to 

 the anal end, which terminates in a rounded point. 



The plates composing the test of this genus are, as in Simtagocijstis, far less elongate 

 than in Pourtalesia proper. The plates of the actinal surface, especially those of the 

 lateral ambulacral and interambulacral zones (PI. XXXV.* fig. 10), being elongate, while 

 on the abactinal surface above the ambitus the coronal plates arc comparatively quite 

 uniform in size (PI. XXX V.'^ figs. 9, 11, 12) and more or less hexagonal. The arrange- 

 ment of the plates of the abactinal side of the test in Spatagocystis is very similar to that 

 oi Echinocrepis, but in the former genus there is a more marked difi"erence between the size 

 of the plates of the anterior and posterior halves of the ambulacral and interambulacral 

 zone than we find in the latter genus. The elongated plates forming the actinal surface of 

 these genera do not seem to be due to the more elongate shape of the test, for in genera with 

 a more circular outline the plates may be fully as elongate as in Echinocrepis and Spata- 

 gocystis. This elongation of the actinal plates seems mainly due to the greater or less eccen- 

 tricity of the actinostome, and is not always an indication of the more Spatangoid affinity 

 of the genera, where we find the test on the abactinal surface as in Sixitagocystis and espe- 

 cially in Cystechinus composed of coronal plates, showing but slight dift'erences in size in 

 adjoining ambulacral and interambulacral areas, which are thus made up of very nearly the 

 same number of plates, somewhat as we find it among some of the genera of Ananchytidse. 



The anal snout has completely disappeared, the only remnant of it is the very slight 

 rounded actinal keel extending from the actinal edge of the anal system a short distance 

 towards the actinostome. The rounded arched and somewhat projecting posterior ex- 

 tremity of Spatagocystis, forming a sort of hood over the anal system, is barely rej^re- 

 sented by the slightly projecting extremity of the median interambulacral area of the test 

 immediately alcove the anal system. With the disappearance of the anal snout the sub- 

 anal fasciole has also vanished, and the miliaries, the last of this fasciole, are merely 

 somewhat crowded on the actinal edge of the anal system, and on the edge of the plates 

 of the test forming the outline of tlie anal system. 



