80 BIOLOGICAL SURVEY OF 



p. 363) states that in old individuals of Chalina oculata the 

 amount of spongin increases progressively as you proceed 

 from the base, where the spongin fibers are "very strongly 

 developed," to the tip where the amount of spongin 'Svas so 

 small as to render it doubtful whether the section represented 

 a Chalina.''^ In individuals from exposed situations, the 

 same author reports that the spongin fibers are generally 

 stouter, though not richer in spicules, than sponges from 

 sheltered localities. Furthermore, the development of spon- 

 gin seems to depend somewhat upon temperature, as sponges 

 with horny fibers are much more abundant in tropical and 

 subtropical seas than in temperate or frigid areas. 



The classification of the sponges, comprising the Monax- 

 onida, depends essentially upon the spicules, of which there 

 are two categories — megascleres and microscleres. The shape 

 of the spicules is quite constant and seems to be independent 

 of external conditions and age. In the Homorrhaphidae, 

 where only smooth oxeote spicules and no microseleres are 

 present, the only distinguishing features between genera and 

 species are the relative sizes and proportions of the spicules 

 and, in some cases, their arrangement in the skeleton. Judg- 

 ment must be exercised in employing this feature, as there 

 is often considerable variation in size within a single species. 

 The length of the spicules in Halichondria panicea varies in 

 different individuals from 0.35 mm. to 1.0 mm. Both Hali- 

 chondria and Reniera possess oxeote spicules, but usually a 

 glance is sufficient to distinguish the long, gradually tapering 

 spicules of EalicJiondria from the short, more robust, sharply 

 pointed spicules of Reniera. Forms with spicules intermedi- 

 ate between these conditions are known, and their determina- 

 tion, in the absence of other characters, is necessarily 

 arbitrary. However, this difficulty was not encountered in the 

 sponges of this region. 



The microseleres, when present, are of greater diagnostic 

 value than the megascleres, as they are very constant in 

 their shapes and sizes. Ridley and Dendy suggest that, in all 

 probability, they are not subjected to any modification by 



