THE SPONGES. 83 



opposite surfaces of the sponge. On the one surface the free margins 

 of the reticulate plates may be seen, the surface network here being 

 formed by these margins and by beams which extend between them. 

 On the opposite surface the reticulate plates are covered in by an 

 "evenly expanded fibrous network" (Schulze, 1887, p. 331). 



The sponges had probably long been dead when taken, for the axial 

 canals are wide and distinct, making it easy to observe the various ways 

 in which the hexacts are combined to form the dictyonal skeleton. It 

 may thus be seen that two parallel rays of different hexacts may fuse to 

 form a beam ; or the beam is formed of a single ray, which at its tip fuses 

 with the side of the ray of another hexact, or crosses this ray, fusion 

 occurring at the point of crossing. Far less common is the case where 

 two or even three rays of the same hexact are so curved as to lie nearly 

 parallel to one another and to fuse. Such a condition is found occasionally 

 in the thicker beams at or near the surface. 



HexacUnella veMabrmn F. E. Sch. has hitherto only been recorded from 

 Japan. 



Hexactinella tubulosa r. E. Sch. (sp.?). 



1887- Hexactinella tubulosa Schulze, 1887, p. 328, Plate XCIII. 



At Station 3406 a small fragment was obtained, including only the 

 dictyonal framework. The fragment, which is about 30 mm. long, is 

 tubular, and represents the end of a cylindrical branch about 10 mm. in 

 diameter. The skeletal plates are separated by an interval of about 

 2 mm. The resemblance of the fragment to a Japanese specimen of 

 H. tiihidosa in the Museum fiir Naturkunde in Berlin is so close as to 

 make it exceedingly probable that the sponge belongs to this species. 

 Through the kindness of Professor Wilhelm Weltner I was allowed to 

 make preparations of the Japanese specimen, which I find is undoubtedly 

 referable to Schulze's species, although the anastomosis between the skel- 

 etal lamellae is feebler than in the specimens described by Schulze (1887), 

 and the lamellae are distinct at the dermal as well as at the gastral 

 surface. 



At Stations 3880 and 3406 four other macerated fragments were 

 obtained, including only the dictyonal framework. These represent some 

 tubular species of Hexactinella, probably H. tubulosa. The diameter of the 



