444 JET. L. Hawkins — Tuberculation of Holectypoida. 



than I have yet been able to procure is necessary before the tubercula- 

 tion can be made a reliable specific character. There is, however, 

 in this branch of study, more possibility than in most other aspects 

 of the Echinoidea, of tracing recapitulation during individual develop- 

 ment. When A. Agassiz (in the Revision of the Echini) and others 

 have shown that in, for example, a post-larval JEchinolampas, each 

 interambulacral plate bears a solitary tubercle (so passing through 

 a ' Cidaris '-stage'), it is at least probable that a study of the 

 tuberculation of young stages in the less highly specialized Holec- 

 typoida will afford a useful guide to a determination of relationship. 

 In this paper the feature is treated from a generic rather than 

 a specific standpoint. After a description of characteristic types 

 has been given, some general deductions as to the phylogenetic 

 meaning of the. various characters will be indicated. 



II. THE TUBEECULATION IN CHAEACTEKISTIC SPECIES. 

 1. PYGASTEE SEMISULCATCTS (Phill.). 



(a) The Structure of the Tubercles. 1 (Pig. 1, A.) 



The primary tubercles are small on the adapical surface, prominent, 



but still small, on the ambitus, and but little larger adorally. The 



greater breadth of the scrobicules in the last-named region makes the 



tubercles there appear considerably larger than they are in reality. 



FlG. 1. Diagram showing the comparative structure of the tubercles (in plan 

 and section) in the Holectypoida. A, Pygaster semisulcatus ; B, P. umbrella; 

 C, P. 'morrisi'; D, Holectypus hemisplmricus ; E, Anorthopygus 

 orbicularis; F , Discoidea cylindrica ; G, Conulus albogalerus; H, plate 17 

 from an Holectypus depressus (and section along line a-b) , showing socketed 

 tubercle. [The tubercles represented (except in H) are all from the ambitus.] 



Adapically the tubercles are smaller on the interradial than on the 

 adradial tracts. The scrobicule is circular, very shallow in adapical 

 tubercles, more deeply excavate in those of the adoral surface, where 

 the scrobicules of adjacent tubercles are often confluent (or separated 

 by a low, smooth ridge) transversely, but never vertically. The 

 scrolicular ring is very irregular and indistinct. The basal terrace is 



1 The terminology employed is that standardized by Bather in his monograph 

 on the Triassic Ecbinoderms of Balcony, 1909, p. 61. 



