Glassificatory Position of Hemiaster elongatiis. 239 



the genera under consideration, it is only necessary to compare 

 the forms one with another. 



We may admit that the two species of Hemiaster differ 

 equally from Paheostoma. 



In all the forms the apical system is small, and two of the 

 generative plates, 1 and 4, have conical eminences, on which 

 are the large ovarial pores. 



In all three forms the generative or basal plates 2 and 3 are 

 small and without perforations for ovarial ducts, and here the 

 structural similarity ceases. The genus Paheostoma has no 

 madrepovite passing back between the two perforated plates 1 

 and 4 ; the other two forms have this, and, moreover, the 

 raadreporite reaches back between the posterior oculars I. and 

 V. There are no sutures in Paheostoma, but they are present 

 between all the plates in the forms we have called Hemiaster. 



The peristome of Paheostoma is pentagonal at all ages ; 

 that of our Hemiaster s is never pentagonal, and is of the 

 crescentic shape of the genus, the front edge being curved 

 widely and the back lip being also curved and projecting 

 downwards as a prominent structure. In Paheostoma the 

 interradial plates 7 at the margin of the mouth are large, 

 nearly equal in their considerable space, and the ambulacral 

 plates contribute but slightly to the margin. In the other 

 forms only the plate / of interradium 5 is large, and, as in all 

 Hemiastersj the other first plates barely contribute to the 

 peristome, whilst the greater part of the margin is made up of 

 the ambulacral plates. The first interradial plates of 1, 2, 3 

 and 4j not very wide posteriorly in Palceostomay are decidedly 

 wide in the Hemiasters. 



There is a most remarkable heteronomia of the interradium 

 1 in Paheostoma mirabih, but there is nothing of the kind in 

 the two other forms. They present the normal heteronomia 

 of the interradia, actinally, which is usual to such Spatan- 

 goids. There is no plastron in Paheostoma, but there is in 

 the other forms. 



The very broad ambulacra on either side of the odd inter- 

 radium in Pala^osfoma, are not characteristic of the Hemiasters • 

 on the contrary, their ambulacra are narrow. In Paheostoma 

 there is a remarkable enlargement of the fifth ambulacral 

 plate of amb. I. row a, and it is pushed to the left because 

 plate h3 oi interradium 5 is placed so much to the front of 

 a 5 of the same interradium that tiiere is a vacant space. 

 This is not the case in the two species which we have named 

 Hemiaster, and plates a 3 and h 3 oi the odd interradium are 

 as is usual in the Spatangoids. 



We consider that we have proved that the generic characters 



