648 



REVISION OP THE AXINELLID^, iil, 



oval areas, — up to 200/y. in diameter and .situated at an average 

 distance apart of about 150/x, — distinguished from the intervening 

 portions of the membrane by their relative transjDarency due to 



the fewness of the deeply-staining gran- 

 ular cells occurring within their limits 

 (PI. xliv., fig. 3). In life, presumably, 

 each such area is the site of a single 

 pore (or possibly of several pores). 



Skeleton. — In most respects the skele- 

 ton closely resembles that of R. typica, — 

 and, indeed, as seen in section, is scarcely 

 distinguishable therefrom; the character 

 of the skeletal fibres is exactly the same 

 in both. The chief difference consists 

 in the somewhat fewer main fibres in the 

 present species, and the far greater num- 

 ber of the connecting fibres (except in 

 the peripheral parts of the skeleton), — 

 in consequence of which the pattern is 

 moi'e generally reticulate, and, except 

 towards the surface, much more irregular 

 (PI. xliii., figs.l, 2). The difference in 

 skeletal pattern of the two species is 

 much more clearly marked in very thick 

 sections of their skeleton, as will be seen 

 from a comparison of figs. 3 and 5 in 

 PI. xxxix.; and from these figures it 

 will be obsei'ved also that, in the present 

 species, the skeleton is on the whole more 

 scanty, and characterised by more ex- 

 tensive discontinuities due to the passage 

 of main excurrent canals. As in R. 

 typica, the trichites occur both in dragmata and scattered singh^, 

 and the longer ones (in part) give rise to fibres; the scattered 



* R}iaphoxya(t) pallida. Megascleres. 

 the same more highly magnified. 



Showing also the extremities of 



