ABT.4 SPONGES OF CALIFORNIA de LAUBENFELS 87 



Family HYMEDESMIIDAE Topsent 



Genus HYMENAMPHIASTRA de Laubenfels 



Hynienwrnphittstra may be defined as differing from Hymetrochota 

 in having unsymmetrical desmalike amphiasters instead of symmet- 

 rical birotulates. Genotype and only species: Hymenmn'phiastra 

 cyanocrypta. 



HYMENAMPHIASTRA CYANOCRYPTA de Laubenfels 



Hlfmenamphiastra cyanocrypta db Laubenfe-ls, 1930, p. 27. 



Holotype.—V.^.'^M. No. 21455 ; B.M. No. 29.8.22.18. 



Type locality. — Point Pinos, Pacific Grove, Calif., July, 1925, col- 

 lected by me. In Jul}^, 1929, I searched for this species, but did not 

 find it. Several other investigators, however, discovered it, and ac- 

 cording to their description it was in the same locality as in 1925, 

 one sufficiently difficult to locate that I had missed it. Judged from 

 their remarks, the colony had more than doubled in size during the 

 four years" elapsing. This species has an unusual habitat. There 

 are numerous rounded granite bowlders piled one above the other, 

 the top layer exposed only at very low tides. This top layer bears 

 abundant life, but if one lifts many stones it is observed that the 

 deeper ones are bare of life ; they are, of course, in the dark. It was 

 while investigating to see if any life at all occurred in these depths 

 that this species was discovered. Though in 1925 it was found only 

 about 60 cm below low-tide mark, in 1929, by enlarging its area, 

 it had come up within 20 cm of that point. 



Description. — Shape, encrusting. Size, less than 1 mm thick but 

 spreading laterally from stone to stone so that probably more than 

 a square meter was covered. Consistency, mediocre. Color in life^ 

 rich dark cobalt blue, growing gradttally paler and paler in alcohol ; 

 after four years much blue remains. Osctiles, not evident. Pores, 

 at least IO/a in diameter; perhaps larger when expanded. Surface, 

 superficially velvety. 



Ectosomal specialization, a dermal membrane only some T^u, thick, 

 not easily detachable, fleshy, containing very few spicules. There 

 are extensive subdermal spaces of great variation in size, some more 

 than 100/x deep. Endosomal structure, at the base of the sponge, 

 where it is in contact with the substratum, there is a thin layer of 

 spongin, at least lO/i, thick, perhaps much more. Many of the 

 acanthostyles have their heads embedded in this, very nearly per- 

 fectly perpendicular to the substratum. Above this is a zone where 

 some of the acanthostyles are strewn in confusion. Above this, near 



