184 PROCEEDINGS OF THE ACADEMY OF [1887. 



8. Spicules ofgemmulsevery short, trapezoidal. Gemmule adherent, 



elliptical ; aperture terminal. 



S. navicella. 

 1). Capsule around the gemmule, and chitinous body both spiculif- 



erous. S. bomhayensvi. 



10. Shafts of gemmule spicules smooth; heads composed of numer- 

 ous short blunt or subacute spines. S. bofryoides. 



11. Gemmule spicules spined, particularly near the head. 



S. sceptrioides. 



^ 2. Color cinereous. S. cinerea. 



1 3. Gemmules in layers or groups ; apertures upward or outward; 



surrounded by a cellular parenchyma, charged with subcylin- 



drical, spined spicules. (PL Y fig. ii.) (PI. VIII figs, i to iv.) 



S. fragilis. 

 1 -4. Gemmules in hemispherical groups ; apertures inward ; sur- 

 rounded by a parenchyma of unequal cells, charged with 

 coarsely spined spicules, nearly as long as the less strongly 

 spiniferous skeleton spicules. (PI. V, fig. iii.) (PI. VIII, fig. v.) 



S. iglovifonnis. 



15. Resembling the above, but with spines more broadly conical, 

 etc. S- mackayi. 



16. Parasitic on S. nite/m, with minute, curved dermal birotulate. 



S' bohmii. 



17. Gemmules very large, chitinous coat thin ; crust absent or incon- 

 spicuous ; gemmule spicules smooth, or irregularly furnished with 

 very long spines, frequently located near the extremities. 

 Numerous dermal birotulates. S. novce terra'.. 



(a) Sponge '^ branched. 



(1) Spongilla aspinosa, Potts. (PI. VIII, fig. vi.) Proc. Acad. Nat. Sci. Phila. Nor. 

 1880, p. 857 ete. 



Sponge green, encrusting, thin; upon a relatively thick basal 

 )nembrane; thence sending out numerous radiating, long, slender, 

 cylindrical branches, occasionally subdividing: texture very loose ; 

 surface rather smooth, pores conspicuous. 



Gemmules very few in scattered bunches of ten to twenty or 

 more, small, spherical, with a granular crust, surrounded by an 

 irregular mass of spicules, resembling those of the skeleton. 



"Spongilla arnchnoidea^'' named by H. James-Clark (Am. Journ. Sci. 1871 

 p. 426), and "Siphydora echinoides" by the same author ("Mind in Nature" p. 

 41, 1865) are not accompanied by such descriptions as will enable me to classify 

 them intelligently. 



