190 pnor. p. matctin duncan on the 



two pairs of pores and as many incomplete plates can be seen 

 to form the foundation and much of the ascending part of each 

 process (figs. 11 & 14). 



These plates have distinct sutural lines (under benzene) between 

 them and at the median line ; but their outer or ambulacro-inter- 

 radial sutures are not seen, and the plates therefore merge into 

 the general mass of the process on their side remote from the 

 median line (fig. 14). The direction of the plates is very 

 oblique. 



Next to these plates, towards the radial end of tlie ambula- 

 crum, are broader ones (JN'os. 5 and 6, fig. 14), which have their 

 interradial sutures visible and in contact with the more or less 

 vertical suture of the ridge and process, as it merges into the 

 common ambulacro-interradial suture. These do not add to the 

 bulk of the process. 



The union of the plates at the base of the processes is too 

 decided to admit of separation, and it appears, therefore, that 

 the processes are the result of the combined growth of the whole 

 of the poriferous parts and some portions of the interporiferous 

 zones of the first four or five ambulacral plates. 



The possibility of the upper part of the processes being a 

 structure superadded to the ambulacral plates arises from the 

 fact that fracture occurs very readily between the middle of the 

 process and the top, and along an oblique line from one side 

 downwards and inwards towards the median line of the ambu- 

 lacrum. No other line of ready fracture occurs, and the surface 

 of the fracture is plain and smooth. Nevertheless the use of 

 benzene does not distinguish any line of suture or of union at 

 the part. 



Temnopleitrus toreumaticus, Agass. — Small specimens were ex- 

 amined in the first instance, and their perignathic girdle appears 

 to be smaller comparatively than that of Salmacis, the processes 

 are not so broad, are more delicate, and the small opening is 

 oval and not sharply angular in outline superiorly, as is the case 

 in Salmacis. 



The ridges are not high in relation to the height of the pro- 

 cesses, and they are broad. The upper edge of a ridge is thin 

 and bent, with a bold downward curve, and there is no projec- 

 tion. The grooves in continuation with the cuts for the branchiae 

 are not very pronounced, but the projection inwards of the usual 

 peristomial swelling immediately over the tubercles is decided. 



