284 SIDNEY F. HAlJiAIKR. 



straight nearly parallel sides (fig. 44), or it may be short and 

 broad, intermediate conditions also occurring. The number 

 of teeth varies from twenty to twenty-four (not counting the 

 minute intermediate teeth), the proximal six or seven on each 

 side being very small. 



The A opercula (fig. 45) are even more distinctive, being 

 characterised by the possession of 26 — 33 minute submarginal 

 teeth, which absolutely distinguish this species from all others 

 at present known. The main sclerite has the form of a pointed 

 or Gothic arch, this feature being well shoAvn in the figures 

 given by Busk (B. M. Cat.) and Srnitt, and there is a distinct 

 basal sclerite, which is thinner medianly. 



The following points may be noted about the specimens 

 from other localities. 



One of the Jamaica specimens (c) agrees well with the 

 type, except that the cryptocyst is angular and thickened at 

 the beginning of its descent. Most of the b opercula are 

 long and parallel- sided. One abnormally small b has only 

 thirteen teeth. Another Jamaica specimen is remarkable for 

 the great size and strength of the opercula of both kinds (sec 

 measurements below). The a opercula (fig. 46), which vary 

 a good deal iu their relative length and in their outline, have 

 unusually large distal teeth, borne on a distinct submarginal 

 sclerite. The main sclerite is a good deal expanded at its 

 anterior end. If this process were to continue, the sclerite 

 might meet the submarginal sclerite and give rise to a form 

 of operculum essentially similar to b, although it will be 

 noticed that the proximal ends of the main sclerite are of the 

 A form. The b opercula of this colony are rare, and one or 

 two are of gigantic size, though not differing in any essential 

 detail from the type. The flanges of the sides of the zooecia 

 which roof the lateral recesses give off a process distally 

 which meets the lateral border of the median process ; some- 

 times the opesiule thus cut off is subdivided by a second 

 similar process, thus approaching the condition found 

 in S. connexa, although differing from that species in 

 that the formation of the opesiules is not due to out- 



