ART. 4 



SPONGES OF CALIFORNIA de LAUBENFELS 



73 



macerated skeleton of a dictyonine hexactinellid sponge, and was in 

 intimate contact along one edge with Lissodendoryx kyma (which 

 see). 



Description. — Shape, encrusting. Size, 5 by 18 mm in area; the 

 depth can not be measured, as the sponge penetrated into the macer- 

 ated dictyonine skeleton on which it was growing. Consistency, 

 fragile. Color in life and when dry, pale drab. Oscules and pores, 

 not evident. Surface, superficially smooth. 



Ectosomal specialization, a dermal membrane; fleshy, detachable, 

 about 30yu, to 70^ thick, containing abundant macrosigmas. Endoso- 

 mal structure, the endosome is so blended with the hexactinellid 

 skeleton that its proper structure can not be ascertained. 



Principal spicules, smooth styles (fig. 39, J.) ; size, lOju, by 330)U to 

 15/A by GOOju. First microscleres, palmate isochelas (fig. 39, D) \ 

 length, ITju. to 20/x. Second microscleres, giant sigmas or macrosigmas 



B C 



Figure 39. — Wi7sa hymena de Laubenfels, X300 



(fig. 39, 5) ; length, about 250/^. Third microscleres, sigmas (fig. 39, 

 C) : length, 55/x to 75/x. Fourth microscleres, forceps (fig. 39, E) ; 

 length, lO/i to 12/x. 



Remarks. — Practically all that can be obtained for this interest- 

 ing sponge is of the nature of ectosome. It contains some tornotes 

 and a very few arcuate chelas, obviously from its neighbor L. kyrna. 

 It is packed with the macrosigmas, has but few of the microsigmas, 

 moderate quantities of the palmate isochelas, and rather numerous 

 forceps. These latter appear smooth with less resolving power than 

 €.95 numerical aperture, and even with the utmost care only the very 

 faintest traces of spination can be discovered, but very small spines 

 are certainly present. The deeper portions contain a few scattered 

 styles as mentioned, but no scheme of their arrangement could be 

 discovered. 



The closest relative of liyniiena is clearly Lundbeck's Esperiops-is 

 forcipula (1905, p. 17), which should also be placed in the genus 

 Wilsa. Its styles are 540/i to 680/x; its macrosigmas only 83/i; its 



