470 MR. HERBERT L. HAWKINS ON 



others, must be regarded as a curiously specialized offalioot from 

 the Fygaster-stock, which only survived a short time, and left no 

 de^scenclants. 



2. 77ie Amhulacral Plates. 



Recently, in the 'Geological Magazine' (Hawkins, 62), I indicated 

 in outline the principles of amhulacral structure which charac- 

 terize the Holectypoida. Later (66) I extended the line of 

 enquiry to the other Jui'assic groups of Irregular Echinoids, and 

 showed the influence that plate- structure exerts on the features of 

 the phyllode. It will, therefore, be necessary only to summarize 

 the results of those studies here, for the sake of completeness. 



All the Holectypoida show a crushing together of the priinaries 

 to form compound plates in their ambulacra. The degree of 

 crushing is a progressive one. Most of the ambulacrum of 

 a Pygaster is composed of primaries, while hardly any unmodified 

 primaries remain in the ambulacrum of a Comdus. The building 

 of the compound plates is carried out on a perfectly uniform plan, 

 three original plates going to form one compound plate. The 

 significance of this triple arrangement will be discussed at the 

 beginning of section V". of this paper. Conulus differs from all the 

 other genera of the order (except the little-known Discoholectypus), 

 partly in the early stage at which the crushing commences, 

 and partly in the fact that two out of the three plates concerned 

 retain their primary character (though modified in shape) for 

 a considerable distance beyond the first crushing point, often 

 right down to the ambitus. Two genera may be cited, re- 

 presenting two widely divergent groups, which show an exactly 

 similar amhulacral structure. These are Pyrina, of the Echino- 

 neidfe, and Amblypygus, oi the Echinolampidse. It is hard to 

 believe that so peculiar a structure can have been evolved four 

 times independently. 



Although there is no true phyllode-structure (nor appearance) 

 developed in the adoral parts of the ambulacra in any of the 

 Holectypoida, the nature of their plate-crushing inevitably 

 results in a "hypophyllodal" character (see Hawkins, 66) of that 

 region. It is not until Conulus is reached, however, that the 

 displacement of the plates drives the pore-pairs into a definitely 

 triserial order. In Pygaster the pores hardly deviate from a 

 straight line throughout the length of the ambulacrum, and no 

 regular displacement can be traced in the poriferous zones 

 of Holectypus. In Discoidea the pore-pairs become appreciably 

 triserial midway between the ambitus and the peristome, but 

 recover their linear arrangement before the peristome is reached. 

 In Comdus a triserial character appears practically at the ambitus, 

 and becomes more pronounced as the ambulacrum is traced 

 towards the mouth ; until, near the peristome, the triads are 

 inclined at an angle of 45 degrees to the direction of the radius. 



The greatest difficulty that appears when an attempt is made 

 to trace a phylogenetic seqiience fi-om the Holectypoida to the 



