540 HYATT'S REVISION OF THE 



The resemblance to Dysidea in the skeleton is so close that only one characteristic separates 

 it, the exclusion of the foreign matter from the secondary fibres. 



Loc, Phillip's Island, Austi-alia, in Society Coll. 



Variety Floridiensis. Two specimens of dendritic forms exist in the Coll. Mus. Comp. 

 Zoology, which can be referred only to this species. The fibres are not so much loaded 

 with debris, but in other respects they resemble the oldest, densest and coarsest parts of 

 the Australian form. 



PHYLLOSPONGIADiE. 



This family contains sponges which, instead of growing iip into solid heads, as is habit- 

 ually the form of the Spongiadaa, are apt to form leaf-like or frondose stocks. These are 

 more or less complicated in pattern, owing to the anastomosis of the fronds wherever they 

 come in contact, and occasionally head-like colonies are thus produced. These, however, 

 retain the frond, or leafy aspect, in the arrangement of the internal parts of the skeleton, 

 as in C. vermifera. Many of the species are perfoi'ated, but this is a characteristic not 3'et 

 understood, and therefore of doubtful value. The fibres of the skeleton usually radiate 

 outwardly from one or both sides of a central mass or layer of fibres which have an irreg- 

 ular mesh without pronounced primary fibres. Fibres may be either very regular, as in 

 Spongelia, or very irregular, or absent altogether. Otherwise they are precisely similar 

 to those of the Spongiadaj, though not ordinarily so elastic. They differ from those of 

 the Hirciniada3 in these characteristics and in the absence of foreign matter from the sec- 

 ondary or connecting fibres. 



Caeteriospongia', gen. nov.^ 



The species which represent this genus are so remarkably different from all others 

 that I have ventured to consider them as distinct. The form of the young in the species 

 described is probably flabellate, but speedily becomes irregular, with numerous flabellate 

 bi'anches, which still farther complicate the aspect of the specimen by anastomosing 

 among themselves. The aspect of the whole is thus rendered almost like that of a truly 

 solid head, but on examination one sees the wide irregular cells between the network of 

 flat branches and the leaf-like aspect of the branches themselves. The fibres are of medium 

 size, like those of the Spongiae, but are arranged externally like those of Spongelia, and 

 with considerable regularity. Spongia lameUosa Esper may possibly belong to this genus. 

 The microscopical structure is very similar in Ehler's preparation, and the external aspect 

 of Esper's figure quite like that of Carteriosjjonrjia perforata, with the exception of the 

 holes piercing the fronds.^ 



A difference in the mode of growth would ordinarily be of small taxonomic value, but 

 one such as is described above, and in combination with peculiar skeletal characteristics, 

 could hardly be associated in the same genus with the solid, or even branching heads of 

 true Spongia or Spongelia. Phyllospongia alone approximates to this genus, and its true 

 position may possibly be nearer to this remarka])le group than our present information 

 would lead us to suspect. This appears in C. Madagascarensis, which very closely resem- 



^Tbis genus is dwlicateJ to Dr. H. J. Carter of the British ^ The curious form, figured by Esper, and described as 



Museum out of respect for his long continued and successful Spongia penicillala, is also an undoubted species of this 

 labors upon the natural history of the Poriferaj. genus. Loc., India. 



