186 PROr. p. MARTIN DUNCAN ON THE 



at the peristome, and it can be traced readily on the free inner 

 surface of the test at the peristome (fig. 8). 



Each interradium has a tall, forked ridge occupying the whole 

 width of the area ; there is a vertical median line denoting that 

 the ridge is composed of at least two plates, one from each zone, 

 and the flanks of the ridge are produced sideways above, so as to 

 overhang the ambulacra more or less. The direction of the 

 ridges is strictly obliquely ujDwards and outwards, and they are 

 rather tumid low down and more or less concave near their 

 upper edge. 



It is evident that a ridge is composed of an upward growth of 

 the first coronal plates (one plate in zone a and two in zone b), 

 and its base corresponds with the first pair of interradial tubercles 

 of one zone, and of the largest tubercle near the peristome of 

 the other zone. There are no perforations on the flanks of the 

 interradial ridges of the girdle, and, as will be seen further on, 

 the ridges, like those of Dorocidaris, are not homologous to the 

 processes of other Echinoidea. 



The comparatively high inner edge of the first ambulacral 

 plates is very suggestive in relation to the corresjDonding part in 

 Discoidea ; but it does not appear that the height is due in 

 Goniocidaris to anything else than the usual elevation which 

 separates the ambulacral plates one from the other, in vertical 

 succession. The plates are tall and have a transverse elevation 

 on their upper surface. 



Phyllacantlius iviperialis, Lmk. — There are two good specimens 

 of this form in the British Museum, and one shows a most 

 interesting difi^erence from the usual type of the perignathic 

 girdle in the Cidaridse. In one specimen the free edge of the 

 ridges is very deeply notched and the ridge is low at the median 

 line and high at the sides, which overhang the ambulacra con- 

 siderably and with a double curve. In the other the median 

 notch is deep, and the sides of some ridges are so produced over 

 the ambulacra that they either absolutely unite with the opposite 

 ridges or very nearly do so. (Eigs. 5 & 6.) 



In one instance the union over an ambulacrum is so perfect 

 that the idea of a perignathic process cannot but arise in the 

 thoughts of the observer. There would be room in this arch for 

 the usual muscular attachments of a process (fig. 7), as in the 

 Echinidse. 



