472 MR. HERBERT L. HAWKINS ON 



in outline instead of being pentagonal. The plates of the adapical 

 surface are generally much broader than high, but on the adoral 

 surface this difference is lessened. In Holectypus sens, str., the 

 contrast in the height of the plates of the two surfaces results in 

 the presence of very few interambulacrals on the adoral surface. 

 In Pygaster the difference is not so strongly marked. 



The only member of the group in which any striking difference 

 in the appearance of the interambulacral plates themselves occurs 

 is the peculiar genus Coptodisctis. Here, in a form otherwise 

 hai'dly to be distinguished from Ccenholectypus, all the margins 

 of the plates are bevelled, so as to leave deep grooves along the 

 sutures. This feature, which recalls the similar structures 

 in Goniocidaris and the Temnopleuridpe, is restricted to the 

 adapical surface. Whether it is a result, in this case, of a paucity 

 of carbonate of lime in the water, or of some physiological 

 peculiarity, it is impossible to judge. The feature seems to be 

 quite unique among the Irregular Echinoids. 



2. The Primary Tuherdes. 



In their structvire and proportions, the primary tubercles show 

 no more variety, when traced through the group, than do the 

 radioles that they support. The equality in size of those of the 

 adapical and adoral surfaces, which is marked in Pygaster^ becomes 

 gradually replaced by a tendency towards an increase in size of 

 the adoral tubercles, with a corresponding decrease of those of the 

 adapical surface. In Discoidea, especially in D. subuculus, the 

 reduction of the adapical tubercles has proceeded so far that they 

 can hardly be distinguished in size from their attendant miliaries. 

 Apart from a tendency in Conulus for the boss to become wholly 

 convex in side view, and so fill the scrobicule more completely 

 than do the partly concave sides of the boss in Pygaster, there are 

 no changes of importance to be traced in the actual structure of 

 the tubercles. 



In the arrangement of the tuberculation more vai'iation is 

 found, and there becomes manifest a continual tendency towards 

 a progressive increase in its complexity. I have dealt with this 

 chai'acter (Hawkins, 67) in considerable detail, and give here 

 a summary of the results obtained in my recent paper. 



As Saemann and DoUfuss (27) showed, the actual luimber 

 of tubercles present on each interambulacral plate depends largely 

 on the size, that is, on the age, of the individual. In all the 

 Holectypoida there is at least one plate, at each end of the half 

 interradius, which supports a single tubercle. This is obviously 

 a relic of the primitive, unitubereulate character of the plates 

 of the earlier Regular Echinoids. The number of such plates 

 remaining decreases steadily as the group is traced from the 

 Lower Jurassic to the Upper Cretaceous. The median seiies of 

 tubercles persists in an unbroken line from the apex to the 

 peristome, but, except in Pygaster, is not readily distinguishable 



