VIII. Comparison ivith Structure of Asterozoa. 1915 320 



constitute any serious difference. Spines were borne by the cover- 

 plates in Pyrgocystis sardesoni, and reason has been given for supposing 

 that they occurred in some other species of Edrioasteroidea. But 

 since the cover-plates were in themselves sufficient protection for the 

 grooves, there was not the same need for spines as arose in starfish 

 with their more open grooves. 



The Mouth-Frame of a modern starfish consists of five sets of 

 paired radially placed elements and five sets of paired interradially 

 placed elements. The latter, known as the " mouth-angle plates ", 

 are continuous with the adambulacrals, and are generally considered 

 to be serially homologous with them. At any rate, they probably 

 contain adambulacral elements, even if they may have incorporated 

 some other constituent. The radially placed elements are admittedly 

 ambulacrals (floor-plates) slightly modified. Morphologically then 

 all these elements belong to the skeleton of the radial grooves. There 

 is, however, another element, the so-called ' odontophore ', tying in 

 each interradius and abutting on the mouth-angle plates; whatever 

 its ultimate origin, it presents the appearance of an unpaired inter- 

 radial element. 



For the present purpose it is quite unnecessary to enter into the 

 perennial discussion as to the precise homologies of all these plates in 

 Recent Asteroidea and Ophiuroidea. Comparison need be made only 

 with the early Palaeozoic Asterozoa, and here the problem is much 

 simpler. 



As an example of a very primitive mouth-frame Mr. Spencer 

 (1914, p. 30) takes the fossil which he names Eoacth simplex. 1 His 

 drawing (pi. i, fig. 4) shows a simple series of ambulacrals and of 

 adambulacrals. At the proximal end of the groove the ambulacrals 

 diverge, and the series is there terminated on each side by a curved 

 subtriangular plate, which Mr. Spencer designates " mouth-angle 

 plate ". It is, however, clear from his drawing, no less than from 

 the specimen itself, that this plate continues the ambulacral series 

 and not the adambulacral, and this is further emphasized by the fact 

 that the depression for the first podium lies equally on this plate and 

 on the adjacent ambulacral. Further examination of other interradii 

 in the fossil shows that this ambulacral mouth-angle plate was 

 actually overlaid by a paired adambulacral element, though only the 

 empty space that might have been occupied by such a plate is shown 

 in the drawing. It follows from this that the plates marked by 

 Mr. Spencer as A l and Ad^ were really A 2 and Ad 2 , and that the 

 true proximal ambulacral and its corresponding adambulacral had not 

 yet fused to form a mouth-angle plate. The continuity of the mouth- 

 angle plate with the ambulacral series is also well illustrated by 

 Spencer's drawing of those parts in Stenaster obtusus (1914, p. 32, 

 text-fig. 28). 



Turn now to the " series of schemes of buccal armatures of 

 Palaeozoic Ophiurids", published by Professor and Miss Sollas (1912, 

 Phil. Trans., vol. 202, p. 226), and to Spencer's figure of Lapworthura 



1 This specimen, now in the British Museum, regd. E 13154, was No. 657 

 of the G. H. Morton collection, and comes from the Lower Ludlow or Upper 

 Wenlock Beds of Hafod, Llandovery. 



