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



the Cretaceous and of the Tertiary forms, the following only : — Stephanocidaris, 

 Goniocidaris, Diadema, Centrostephanus, Echinothrix, Astropyga, Aspidodiadema, 

 Micropyga, Colobocentrotus, Heterocentrotus, Parasalenia, Pseudoboletia, Echinos- 

 trephus, Pleurechinus, Mic)vcyp>hus, Mespilia, Prionechinus, Evechinus, Peronella, 

 Astridypeus, Rotida, Neolampas, Anochanus, Pcdceotropus, Cionobrissus, Echinocrepis, 

 Spatagocystis, Cystechinus, Argopatagus, Palceostoma, Tripylus, Faorina. 



From our study of the embryonic stages of the Echinidse, the Clypeastridae, and 

 the Spatangid^, and a comparison of these stages with the genera of the Desmosticha and 

 Petalosticha which have either succeeded the genera above mentioned, or have lived with 

 them during the Cretaceous period and have disappeared either during the Cretaceous or 

 the Tertiary periods, we find no difficulty in tracing an unbroken systematic connection from 

 the earliest Cretaceous beds to the present time. But this connection is so complicated, 

 and ramifies in so many directions, that it must be hopeless, even with the small number of 

 species of Echinids known, to attempt to do more than to indicate the lines of affinities, 

 the delicate threads of which we can trace in characteristics of genera which at any 

 special epoch seem to have little or no structural afiinity. Let us take, for instance, the 

 genera characteristic of the Chalk, and attempt to trace their connection both backw^ards 

 and forwards in time. 



Taking these genera in their most extended signification, and more especially those 

 characteristic of the Lower Cretaceous formations, Cidaris, Orthocidaris, Phyllaccmthus, 

 Tetracidaris, Goniopygus, Codiopsis, Magnosia, Cyphosonm, Pseudocidaris, Orthopsis, 

 Pedinopsis, Codechinus, Stomechimts, Acrosalenia, Echinothuria, Pygaster, Discoidea, 

 Holectypus, Pyrina, Clypeopygus, Pyguriis, MetaporUnus, Holaster, Toxaster, and 

 comparing them in the first place with the genera of the Lias as far as they are known, 

 we find that, with the exception of Cidaris and Hypodiadema, the forerunners of the 

 Cidaridse and Diadematidae, not a single form of the Echinidai is represented. 



To attempt to explain then- relationship to the earlier types, we may say in a very 

 general way that the Perischoechinidse early show on the one side a tendency to limit the 

 number of the rows of interambulacral plates ; and on the other side a decided tendency 

 to a splittmg up of the ambulacral and interambulacral plates into numerous irregular 

 rows ; they are thus the only group leading directly to such types as Cidaris on the 

 one side and to the Echinothuridse on the other, the genera Tetracidaris and Echino- 

 thuria in the Chalk being the representatives of these two groups of Pateechinid^ ; while 

 the presence of such a type as Hypodiadema early in the Trias may, perhaps, represent 

 the reduction of the number of coronal plates in some of the earlier Echinids, w^hUe 

 retaining the uniform tuberculation so characteristic of the Palaeechinidas, and retaining 

 at the same time the proportionally broader ambulacral areas of some of the types. 

 From the time of the Trias the Cidaridse have been a most persistent type, and the changes 

 the members of the family have passed through are restricted to very narrow limits, with 



