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



glance appear to indicate very widely differing structural characters. In the same way, 

 widely separated as Archceocidaris and Qligoiwrus appear, the existence of such genera 

 as Lepidocentrus, and PJiolidocidaris, shows that even the characters which at first sight 

 so strongly contrast in Melonites and Archceocidaris may gradually disappear, although 

 undoubtedly the type to which Melonites, Oligoporus, Palceechinus, and the like belong 

 would lead us more directly as far as the structure of the coronal plates is concerned to 

 the abutting plates of the Echinids of the present period, which as I have attempted to 

 show are really all to a certain extent bevelled at the edges. This bevelling depends 

 first uj)on the curve of the test and in the second place upon the thickness of the test, 

 while such decidedly imbricating coronal plates as those occuppng the actinal surface of 

 the test in ArchcBOcidaris and the like, and on which we find also a single large primary 

 tubercle developed, lead us directly to the Echinothuridse, and to such types as Phormo- 

 soma, which are evidently not very difi"erent in their structure from the Silurian ArchcBO- 

 cidaris ; and at the present day the Diadematidse, especially Astroioyga, still retain traces 

 of the existence of more than two rows of interaml:>ulacral plates and of the imbrication of 

 the coronal plates, within the ambulacral and iuteramliulacral areas, so prominently 

 developed in the Echinothuridse of the present day. This is not the only family in 

 which we find interambulacral areas with more than two rows of coronal plates. In the 

 remarkable genus Tetracidaris of Cotteau, the double row of interambulacral plates on 

 each side of the median line is not limited as in Astropyga to a few plates on the actinal 

 sides of the test. In that genus we find only a short part of the interambulacral area 

 near the abactinal part of the test where the normal number of interambulacral plates are 

 present, while in the rest of the test, with the exception of the few plates near the apical 

 system, the interambulacral areas consist of four rows of primary plates. 



The very peculiar splitting of the vertical rows of coronal plates noticed by Quenstedt 

 in Melonites ^ seems to point to some structural peculiarity in the PalseechinidBe such as 

 I have described in the breaking up of a single interambulacral plate in our recent 

 Echinothuridge. It shows, at any rate, what some of the other genera of Palajechinidae 

 plainly show, that we find it impossible to define the number of rows of coronal plates in 

 the test just as we find it impracticable near the apical system of the regular Echinids to 

 ascertain how many rows of interambulacral plates there are present, as they appear in 

 that region of the test packed in as they best can find place and take up tlieir regular 

 and symmetrical arrangement only later, while we may observe that in the Pal£eechinida3 

 this symmetrical arrangement never takes place, the vertical rows of plates running in as 

 best they can, thus forming another important embryonic character of the Palaeechinidse. 



From an embryological and pal^ontological standpoint perhaps no more important 

 view has lieen taken than that of Loven regarding the nature of the apical system of the 

 Echinidse, which he developed from a comparison I had at first indicated of the complete 



' Quenstedt, Pal. Deutsch., 1872 to 1875, vol. iii. p. 381. 



