498 DR. J. K. GixAY ON SPONGES. [May 9, 



The Corticatce into genera thus : — 



I. The cortical layer with stellules. 



Needles uniform, simple Tethya. 



Needles simj)le and anchor-shaped Stelletta. 



II. The cortical layer with ylobules. 



Cortical layer with globules only Camtnus. 



Cortical layer with globules and needles Geodia. 



III. The cortical layer without globules or needles. . Anchorina. 



Dr. Schmidt published a Supplement in 1864, and a second in 

 18GG ; in the latter he examined Dr. Bowerbank's genera in his 

 work on 'British Sponges,' and states the genera to which he would 

 refer them. 



Dr. Schmidt's criticism on Dr. Bowerbank's genera is peculiar. 

 Dr. Schmidt forms genera on a very different system to Dr. Bower- 

 bank, and then refers certain species to the genera which he has 

 himself used, and observes that the other species of Dr. Bowerbank 

 belong to different genera. 



Dr. Bowerbank might with equal fairness have criticised Dr. O. 

 Schmidt's genera on the same system and with equal truth, except 

 that he would not have been able to ])lace the name of Nardo after 

 the generic name ; but this is also equally unfair, for, though the 

 genera may bear Nardo's name after them, they are characterized 

 and restricted by Dr. (). Schmidt. 



MM. P. Duchassaing and G. Michelotti, in their ' Spongiaircs de 

 la mer Cara'ibe,' Harlem, 18(54, 4to, with twenty-five coloured plates, 

 jniblished in the 'Natuurk. Verb. Holland. Maat. Wet. te Harlem,' 

 vol. xxi. 18G4, proposed the following arrangement: — 



Order I. DICTYOSPONGLE. Keratose network furnished with 

 spicules ; fibres forming a network. 



Fam. 1. EuspoNGi.E, Keratose network well developed ; siliceous 

 spicules wanting or very rudimentary. 



Subfam. 1. Penicillatce. Horny fibres forming nerves, pencils, or 

 columns, but are never distinctly separate as in the other sub- 

 families. 



Subfam. 2. Helerorjcnce. Fibres distinct, and of two kinds. 



Subfam. 3. HomogencB. Fibres horny, hollow, very rigid, equal, 

 and anastomosing into meshes, but never uniting into bundles. 



Fam. 2. Lithospongi^e. Keratose network formed by siliceous 

 fibres ; texture decidedly stony. 



Fam. 3. Halyspongi.^. The spicules are siliceous and well de- 

 veloped, predominating over the others. 



Subfam. 1. ArmatcB. Spicules needle-shaped, forming a mesh with 

 others, which are anchor-shaped. 



