REPORT OX THE ECHINOIDEA. 95 



spicules as in the bulk of the Diadematidse. In fact the Ecliinothuridse strongly recall to 

 us the embryonic stages of our regular Echioids {Strongylocentrotus and Arbacia) in which 

 the distinction between the coronal plates and the actinal and abactinal systems does not 

 exist, and in which the whole test is made up of plates of similar structure. The Echino- 

 thurid^ are somewhat more differentiated than the Perischoechinidse, in which the 

 coronal plates themselves are stUl very numerous and not reduced to the typical number 

 of two j)lates for the interambulacral system, as in all the Echinoidea known at the present 

 period; but even in the Echinothuridse we stUl have a trace of this abnormal character of 

 the Perischoechinidse, of having a number of rows of plates in the interambulacral system. 



Some of the species of the genus Phormosoma, in which the test is most flexible, such 

 as Phormosoma tenue, show traces of an irregular sul^division of the coronal plates both 

 on the actinal and abactinal sides (PL XIII. fig. 1, PL XIV. figs. 1, 2, PL XIX.^ fig. 2). 

 Diagonal or transverse lines are seen to run from one plate to the next, so as to sub- 

 divide the primary coronal interambulacral plate into two or three, and sometimes four 

 or five secondary plates ; each one of these secondary plates corresponding usually to a 

 primary or secondar}^ tubercle. This breaking up of the primary plates, of course, gives 

 to the test a much greater mobility than it had before in spite of the extreme tenuity of the 

 test (PL XVIII.^ figs. 4, 5, 7, 8). These secondary plates, although extremely thin round 

 the edges, are strengthened in the centre by a deposition of carbonate of lime forming a 

 circular button in the centre (PL XVIII.'' fig. 4), to strengthen that part of the plate 

 which carries the primary tubercles and spines. This splitting up of the coronal plates 

 into plates corresponding each to a primary tubercle exists also to a certain extent on 

 the actinal surface of Astropyga, though it has not to my knowledge been noticed before. 

 The thickness of the inner and outer fold of the cuticle of the test forms lines more or less 

 coincident with the secondary plates of the interambulacral areas (PL XIV. figs. 1,2; 

 PL XVIII." fig. 7). This gives us, I think, a natural explanation of the structm*e of the 

 coronal interambulacral areas of the PerischoechiQidse, only in this group the splitting up 

 of the primary interamljulacral coronal plates was quite regular, and the lines of sutures 

 are regularly placed as in the amljulacral system. It is remarkal^le that in the Spatan- 

 goids, the Clypeastroids, and all the higher Petalosticha, the arrangement of the plates of 

 the ambulacral system should have remained comparatively simple as well as in the most 

 embryonic group of the Desmosticha, the Cidaridae and Salenidse, while La the Perischoe- 

 chinidse, the Echinothuridse, and by far the larger number of the Desmosticha, the 

 arrangement of the plates of the ambulacral system is quite complicated, and the number 

 of rows of plates across the ambulacral areas greater than that of the interambulacral 

 areas (which are, of course, limited to two in the bulk of the recent Desmosticha). 



Grube and Thomson have already called attention to the similarity in the structure of the 

 teeth of the Echinothuridse and Diadematidae. Thomson has figured the teeth of Phormo- 

 soma placenta (Porcupine Echinids, Trans. Roy. Soc, 1874, pi. Ixiii. figs 9, 9«), and I have 



