150 GEODIA AGASSIZII. 



sterrasters, and euasters more or less similar to those of Gcodia agassizii are 

 G. spherastrella Topsent and the sponges hero described as G. (Cydonium) ja- 

 ponka Sollas, G. mesotriaena , G. mesotriaenella, and G. breviana. Gcodia japonica 

 is distinguished from G. agassizii by being cup shaped, having rountled, lobate 

 protul)orances on the outer surface, and possessing, ahhough very large in size, 

 considerably smaller megasclercs; the large amphioxcs of this species are less 

 than half as thick as those of G. agassizii. Geodia spherastrella is distinguished 

 from G. agassizii by being incrusting and having much larger dermal strongylo- 

 sphaerasters. Whether there are also differences in the size of the megascleres, 

 as is very likely, I cannot say, since Topsent ' gives no measurements. Geodia 

 mesotriaena is more or less meandric and provided with praeoscular canals; and 

 it has choanosomal amphioxes and orthotriaene-cladomes twice as long, and also 

 # larger euasters. Geodia mesotriaenella has more slender and differently shaped 

 anatriaene-clades, no mesoprotriaenes with epirhabd exceeding the clades in 

 length, and somewhat different strongylosphacrasters. Geodia breviana pos- 

 sesses minute dermal anaclades, its large anatriaenes have much thicker clades 

 and its strongylosphaerasters are larger. 



The only species of Sidonops which has orthoplagiotriaenes, subcortical 

 tcloclades, sterrasters, and euasters similar to those of G. agassizii is the sponge 

 described by Dendy ^ as G. areolata Carter. This sponge differs from G. agassizii 

 by the absence of mesoprotriaenes, by having somewhat smaller spicules, particu- 

 larly thinner choanosomal amphioxes and orthoplagiotriaene-rhabdomes, and 

 by the reticulate structure of its surface. The differences in the size of its 

 spicules and the absence of mesoprotriaenes, which, if present only in the fur, 

 have perhaps been lost post mortem, may not be sufficient for distinguishing it 

 specifically from G. agassizii. The reticulate appearance of the surface of G. 

 areolata is caused by the presence of a superficial network, which consists, 

 according to Carter (Ann. mag. nat. hist., 1880, ser. 5, 6, p. 133, pi. 6, fig. 37), 

 of rows of small protruding dermal amphioxes, according to Dendy (loc. cit.), 

 of subdermal pigment-cells. In none of the specimens of G. agassizii could I 

 detect the slightest trace of a network of either of these kinds on the surface. I 

 am not prepared to express an opinion on the systematic value of the presence 

 or absence of this network. Since, however, it may be systematically important ; 

 since there is a difference in the spicules, although by itself perhaps too slight for 

 specific distinction; and since the localities where these sponges occur. Pacific 



' E. Topsent. Spongiaires des Azores. Result. Monaco, 1904, 25, p. 70. 



' A. Dendy. Report on the sponges. Rept. pearl oyster fisheries, 1905, pt. ,3, p. 87. 



