BY E. F. HALLMAXN. 299 



only over a limited portion of the surface of one specimen are the 

 areola^ at all depressed and pit-like. Accordingly, in conveyinc^ 

 the impression that the reticulation of the surface is produced 

 entirely by a "network .... of projecting lines " with "poly- 

 gonal meshes " in which are "depressions about 4 or 5mm. deep," 

 the original description is quite misleading: one can see, indeed, 

 from the figure in the " Catalogue '' (PI. i., fig.3) how free from 

 any pitted appearance is the portion of the surface therein shown. 



The description is inaccurate also in several statements regard- 

 ing the excurrent canal-system. We are told that vents are 

 scattered over the surface and lead into short conic tubes, which 

 are not oscula but praeoscula: that these " short '' (sic) tubes, 

 which in the case of the original specimen are " nine in number 

 and measure 250 mm. long by '20 mm. wide at the mouth," have 

 their walls covered throughout by a reticulation similar to that 

 of the exterior surface; and that proper oscula, 2 to 10 mm. in 

 diameter, are scattered over the whole surface including the sides 

 of the conic tubes. After the most careful examination of the 

 several specimens, I can find no reason to doubt (what, even at 

 first sight, seemed most probable) that all the tubes referred to, 

 including those leading from the so-called oscula, are nothing 

 more than excavations made by crustaceans and other boring 

 organisms, a considerable number of which are still present in 

 most of the tubes: it is significant, also, that many of the smaller 

 tubes are entirely filled with sand and mud. The tubular ex- 

 cavations are everywhere lined with a dense tough rind, often 

 exceeding 1 mm. in thickness, composed almost entirely of closely 

 packed megascleres: on no part of their wall, have T seen any 

 trace of areolation. 



If the soft tissues be removed by means of a macerating agent, 

 there remain (Pl.xv., fig. 2) finally (i.) the rind-like cortical layer 

 forming the outer surface: (ii.)the rind which lines the above- 

 mentioned cavities: and (iii.) extending through the whole interior, 

 a coarse network of somewhat flattened or strap-shaped trabecula?, 

 similarly constituted to the ''fibres" of Spirastrella australis and 

 SpirastreUai}.) ramulosa, which are ordinarily 0-5 mm. to 1 mm. 

 broad, and enclose meshes, on the average, several millimetres in 



