SILICIOUS AND HORNY SPONGES WILSON. 485 



In them all the conuli are higher, the ridges more prominent, and 

 the depressed areas deeper in certain regions than is common. Such 

 regions doubtless represent areas in which contraction had proceded 

 quite far. 



Main fibers of a region nearly parallel to one another, vertical 

 to surf ace,- 40-50 [/. thick, well filled with spicule fragments but con- 

 taining no, or almost no, sand grains. Fibers in the peripheral 

 part of sponge are 900-1.800 jjl apart. As a rule the fiber terminates 

 singly in a conulus, but it may branch at very acute angles, the two 

 or three branches terminating together in a single conulus. Con- 

 nectives are of uniform thickness, about 20 y. thick, and contain no 

 foreign inclusions; characteristic meshes 90, 180, 270 [a in diameter. 



In addition to the above, three small specimens of massive shape, 

 50-60 mm. in diameter, were taken along the shore of Busin harbor, 

 Burias Island. Conuli less than 1 mm. high, and less than 1 mm. 

 apart. Oscula 2-4 mm. in diameter, many sphinctrate. Main fibers 

 in peripheral part of sponge, 700-1,200 pi apart. In the remaining 

 features mentioned above, these sponges agree with the larger ones 

 and are therefore to be regarded as young (see below) forms of the 

 same race. 



Some of the recognized varieties (Lendenfeld, 1889) of Euspongia 

 officinalis are well marked, although perhaps certain of them are 

 only habitus forms. Others are vaguely delimited by quantitative 

 differences which are too slight to be relied upon with an}' great 

 certainty in the practical work of classification. It would probably 

 be a good provisional step in the handling of collections, which do 

 not permit the rich detailed comparisons between many individuals 

 (young and old, in various kinds of habitat, and in different physio- 

 logical states) that come with a residence in the working-field of 

 a laboratory, to reduce the number of varieties, recognizing by name 

 only such as depart conspicuously from a type which itself is con- 

 ceived, in accordance with the facts, as not uniform but heterogen- 

 eous in respect to detail. 



Thus I find that, as far as recorded data allow one to form a judg- 

 ment, the Philippine sponges above described can not be separated 

 from Dendy's variety cei/lonensis from the Gulf of Manaar (Dendy, 

 1905, p. 211) nor, on the other hand, from the Porto Rico specimens 

 which I referred (1902, p. 402) to var. rot inula in the sense of Len- 

 denfeld (1889, p. 269), under which name Lendenfeld combined 

 numerous "West Indian and Florida forms that had been distinguished 

 especially by Hyatt. From these the Mediterranean variety adH- 

 atica is not very distant, although possibly specimens of adriatica 

 might be picked out from the others by the larger average size of the 

 meshes and thickness of the connectives and by the fact that the 

 foreign bodies in the main fibers are chiefly sand grains. 



