126 BRITISH PALEOZOIC ASTEROZOA. 



prominent ridge upon which is frequently borne a row of 

 stout spines. At first sight, especially when tbe groove is 

 open, they might be mistaken for the side-shields of true 

 " Ophiuroids." Really, however, they have quite a different 

 structure (see below). 

 (3) All the plates of the apical surface have usually stout spine-like 

 or ridge-like paxillar projections which carry spines, 

 (c) Characters which show the Origin of the Families. 



(1) The early members of the lineages have an arrangement of the 



plates of the disc which is even more primitive than in the 

 Hudsonasteridse (see p. 129 and Text-fig. 81). 



(2) Infero-marginalia can always be recognised, but there are no 



supero-marginalia. It will be remembered that the forms of 

 Section A had always both series of marginalia, although it 

 was suggested (p. 67) that the supero-marginalia had arisen 

 later than the infero-marginalia. The fact that only infero- 

 marginalia can be recognised in the form now about to be 

 described, supports the argument from the structure of the 

 apical plates, namely, that the Families broke away from the 

 original stock before the typical " Hudsonaster " structure 

 was developed. These infero-marginalia are pushed well 

 over on to the apical surface by the strong development of 

 the adambulacralia. 

 Distribution. — The forms are widely distributed both in space and time. They 

 appear with the earliest known Asterozoa in the Ordovician, and comprise a 

 fair proportion of the forms in the early Palaeozoic rocks. In the Mesozoic they 

 become relatively less important, although one genus, Arthraster, persists into the 

 Chalk. Species have been recognised from the rocks of N. America, England, 

 Scandinavia, Russia, Germany, and Australia. 



Lineage Changes. — It will be noticed in the description below (p. 136) that the 

 forms present lineage changes comparable to those shown by the Asteroidea of 

 Section A (pp. 61 — 65). 



Schuchert, as recently as 1914, appears to have been the first to recognise the 

 essential similarity of structure of many of the various forms which are now about 

 to be described. His super-family, the Urasterellacea (see above, p. 58), includes 

 most of the species described below together with an additional family, that of the 

 Compsasteridas. Personally, however, I do not feel that I know enough about this 

 latter family to undertake its classification for the moment. 



Schuchert classifies the Urasterellacea among the Cryptozonate Asteroidea. 

 Hudson, more recently still, has published a valuable paper in which he remarks 

 that the structure of Urasterella shows that we have here a " new sub-class of 





