528 HYATT'S REVISION OF THE 



development of primfiry fibres. I am informed also by Mr. Goode, that this sponge is 

 apparently very abundant, and is the principal useful sponge found in the lagoons of Ber- 

 muda. It is, however, too coarse for anything but the roughest purposes. 



Loc., Bermuda, in Coll. Wesleyan University, Middletown, Mass., and Soc. Coll. 



Sub-species dentata. 



This form (PI. xvii, figs. 6, 7) is represented by only one specimen, one of the most 

 remarkable species I have yet seen. The whole surface is covered by a hard, even crust, 

 similar to the tough epidermis of Geodia, and resembling it in color. It extends over 

 the whole external surface, both above and below, but is not found in the endoderm 

 lining the canal system. Microscopical examination disclosed the curious fact that it was 

 probably due to the great extent with which the epidermis of the animal while living was 

 loaded with fine calcai'eous mud. The whole figure, which is of an irregular outline about 

 300 mm. in diameter, is only about 40 mm. in the thickest part, and is covered above as in 

 variety typica, by the usual tufts. These, however, are very regular in distribution, of 

 small size, round, and quite pointed. There are also larger oscules sparsely scattered over 

 the surface with crater-like walls of about 5 mm. in height. The figured specimens are 

 parts of a fragment of a large colony in the Collection of the Museum of Comp. Zoology. 



Loc, Havana, in Mus. Comp. Zool. and Soc. Coll. 



Sub-species Mauritiana Hyatt. 



This variety (PL xvii, fig. 5) is remarkably cut up by internal channels which per- 

 meate the substance in every direction. These are quite large, elliptical in outline, and 

 about a quarter of an inch in diameter, and give the sponge a very rough aspect on the 

 surface, which in perfect specimens rises into prominent tufts and ridges between them. 

 The fibres are remarkably inelastic, and the whole mass unyielding to the touch, but not 

 brittle, friable, or easily broken. 



Loc, Mauritius, in Coll. Mus. Comp. Zool. and Soc. Coll. 



Variety Pacifica. This is an incrusting form, permeated by similar large channels 

 which, on account of the thinness of the mass, give the sponge a perforated aspect when 

 held up to the light. The surface is very rough, and is cut uji by parallel ridges and iri'eg- 

 ular shaped elevations, bearing either rows or bunches of projecting fibres. 



Loc, Pacific Ocean, U. S. Ex. Exp., Yale Coll. and Society Coll. 



Variety decidua. This variety, although precisely similar in structure to the variety last 

 described, differs slightly in the aspect of the surface. The projecting ridges are trans- 

 formed into bunches, sometimes solid and sometimes fringed by the projecting free ends of 

 primary fibres. There is a tendency also in parts to extend these into a sort of velum 

 covering over the intermediate spaces. This is particularly the case upon the lower side of 

 the specimens from Mauritius, and a part of the upper surface in the Havana specimen. 



Both specimens are in the Coll. Mus. of Comp. Zool. 



Stelospongos SchmiJt. 



This interesting group, described by Schmidt, is readily distinguished in most species by 

 the compound structure of the primary fibres, but in others tlie j^i'iniary fibres are not so 

 closely associated. The primary fibres are associated in bundles, and closely connectetl by 

 secondary fibres. Sometimes these primary fibres are nearly parallel, sometimes branching 



