1887.] NATURAL SCIENCES OF PHILADELPHIA. 201 



These groups were not observed or described by Dr. Leidy, and 

 when first noticed by myself were supposed to indicate either a 

 new species or variety ; but it was long ago discoverd that to a 

 greater or less extent they formed a characteristic feature of all types 

 of Spong ilia frag Ills. It has seemed to me that while the "pavement 

 layer" of gem mules was firmly attached to its base of support, indi- 

 cating a purpose to reproduce the sponge at the same place, the 

 groups, which are not so attached, and are liberated by the slough- 

 ing away of the skeleton spicules during the winter season, float 

 off and serve to propagate the species in distant localities. They 

 must not be understood to be merely accidental collections of gem- 

 mules in the same neighborhood and without coherence, (as is the 

 case with those of S. aspinosa); but to be closely and j^ermanently 

 associated and embedded in a mass of compact "cell-structure"; (the 

 "cellular crust" of Carter, the "external parenchyma" of Vejdovsky); 

 sometimes but three or four together, (PI. V, fig. ii, B.) compared 

 by H. J. Carter to the tetraspores of Selaginella; sometimes a doz- 

 en or more, very irregularly grouped, but always with their foram- 

 inal tubules projecting outtvard through the crust ; in which vast 

 numbers of spined spicules are embedded. 



The spicules of this class are very variable in shape in the differ- 

 ent localities in which this species has been found ; being long or 

 short ; robust or. slender ; truncate or pointed ; while the general 

 characteristics of the species remain unquestionable. A few vari- 

 eties that appear constant have been named. Upon nearly every 

 slide of prepared gemmules or spicules may be seen a few abnormal, 

 spherical forms, bristling with spines, and reminding one of the pol- 

 len of Mai vaceous plants, or the "caltrops" in sometime military 

 use. 



S.fragilis, (S. lordii, Bowerbank.) Syn. Proc. Zool. Soc. 1863. 



Dr. Bowerbank's description is here copied for convenient refer- 

 ence. The species would have been "good" in his name had it not 

 been antedated twelve years, by Dr. Leidy. 



"Sponge sessile, coating ; surface even, smooth. Oscula? simple, 

 dispersed. Pores inconspicuous. Dermal membrane pellucid, 

 aspiculous. Skeleton spicula acerate. Ovaria congregated on the 

 basal membi-ane, very numerous ; spicula entirely spined, fusiform- 

 cylindrical, dispersed on the surface. Basal membrane abundantly 

 spiculous ; spicula dispersed, same as those of the ovaries. Color 

 ochreous yellow^ to green." 



14 



