BRITISH SPONGIAD^l. 173 



diameter of the base, at others, terminating very abruptly 

 and obtusely. In some of the oldest specimens, two, 

 three, or more of the mammae unite at their bases, forming 

 a stout short column for about half their height, but ter- 

 minating in separate short cones. In the dried state, their 

 surfaces are very much corrugated by collapse of the dermis 

 inward. Their usual height is five or six lines, with a 

 basal diameter of about two lines. The structure of the 

 exteriors of these organs is the same as that of the body 

 of the sponge, each one having two or three large excur- 

 rent canals extending the whole of their length ; I could 

 not detect any terminal oscula, but in several instances I 

 found open oscula on their sides, rather smaller than those 

 dispersed on the body of the sponge. 



There are two species with which this sponge may pos- 

 sibly be confounded. The greenish gray dermis, and deep 

 red interstitial tissues of II consimilis closely resemble the 

 corresponding tissues of II. sanguinca, and the spicula of 

 the skeleton of the former are of the same form as those of 

 the latter species ; but they are rather shorter, and con- 

 siderably stouter than those of II. sanguinea. The dermal 

 membranes of both species are abundantly furnished with 

 irregularly dispersed spicula, but they are much more 

 numerous in H. consimilis. These comparatively slight 

 differences would scarcely have led us to have considered 

 them as distinct species, if it were not that the striking 

 differences in form are such as to at once claim our atten- 

 tion. I have seen many species of H. sanguinea, and have 

 never observed the slightest indication of the production of 

 the irregular mammaeform organs which are so abundant on 

 all the six specimens of //. consimilis. These strictly external 

 characters, therefore, become of greater value in this case 

 than is usually attributed to them ; and in addition to the 

 structural differences, slight as they are, after a careful ex- 

 amination, they appear to fully warrant the separation of 

 the two species. The reverse of the case which I have 

 just described exists in the comparison of //. consimilis and 

 H. mammeata. Here we have the massive sessile forms 

 and mammseated surface of each species closely resembling 



