ART. 8 A NEW VARIETY OF SPONGE WILSON AND PENNEY 5 



typically (that is, nearly always) smooth, although occasionally it is 

 feebly spinose, noticeably shorter than any of the other rays, tapering 

 to a point or sometimes to a rounded point ; common range in length 

 175(0, to 220/x, but the ray is occasionally as short as 80;u, or as long as 

 250;Li. This ray when exceptionally short does not taper and is 

 terminally rounded or dilated. Thus the autogastralia differs greatly 

 from those recorded by Lambe and in less degree, but yet qualita- 

 tively, from those of Schulze's specimens. 



There are no hypogastralia. And yet in radial sections we have 

 seen inequiended, diacts. five or six in a section 10 mm. wide, 

 arranged radially to the gastral surface as the parenchymal rays of 

 the hypodermal pentacts are arranged radially to the dermal sur- 

 face. The larger end of the diact is directed toward the gastral 

 surface and the axial cross is very much nearer this extremity than 

 it is to the more attenuated parenchymal end, the lengths of the 

 two rays in a typical spicule being as one to three. These facts give 

 some ground for regarding the inequiended diacts as vestigial hypo- 

 gastralia. Hypogastral pentacts are absent in most members of this 

 family, the Rossellidae (Schulze 1897, p. 13). The inequiended 

 diacts observed in position were without comitalia; about 2 mm. 

 long, 45ja thick at middle of spicule, tapering toward both ends, 

 minutely spinose at both ends, which were sharp or rounded. In 

 macerations longer spicules of the same kind were observed, the 

 length varying up to 7 mm. At the site of the axial cross there 

 may or may not be conspicuous bosses. 



The oxyhexasters^ varying to the hexactine shape (the interme- 

 diate forms being sometimes known as hemioxyhexasters) of the 

 Alaskan sponge agree well enough with those of the type. For the 

 latter (Lambe and Schulze) the diameter is recorded as 60/* to lOOju,; 

 principal rays smooth and very short; terminals 2-3, long, smooth, 

 or feebly roughened. The spicules of the Alaskan sponge reach 

 a somewhat larger size, 90/a to 144;a diameter. 



The discoctasters of the Alaskan sponge are like those of the type. 

 In the latter (Lambe and Schulze) they are set down as having 

 a diameter 60ja to 100/x; the (eight) main rays 20,0, long; terminals 

 6-10 in number and 20/x to 30ja long, moderately divergent, the termi- 

 nal knob (disk?) very small. The diameter of the spicule in the 

 Alaskan sponge is commonly about 92jii. The strikingly small size 

 and the shape of these spicules are regarded by Schulze (1897, p. 13) 

 as constituting the most important of the species-characters. 



The TTiicrodiscohexasters recorded by Schulze were not observed 

 in the Alaskan sponge. The specimen to be sure was a dried one 

 and these very small spicules may have been lost. On the other 

 hand the parenchymal tissue has been so remarkably preserved 



