REPORT ON THE MONAXONIDA. xliii 



in size or form. Such appears to be the case in Halichondria panicea, Myxilla nohilis, 

 &c., and in many cases it is extremely difficult to say where the subdermal cavities end 

 and the inhaleut canals or lacunae jDroper begin. 



In Phakellia and m Esperiopsis challengeri (both of which, it will be remembered, are 

 fiabellate sponges with the pores on one surface and the oscula on the other), we find a 

 somewhat peculiar condition. The pores lead directly into large, irregularly conical 

 chambers (vide PI. XLIX. fig. 3), which often e^ftend almost through the entire thickness of 

 the sponge, and from these great chambers, which we may regard as the subdermal cavities 

 in that they are in direct communication with the exterior, the system of small inhalent 

 lacunse takes its origin. These large, vertically elongated subdermal cavities interdigitate 

 with similar exhalent chambers opening on to the opposite surface {vide PI. XLIX. fig. 3). 



In those cases where, owing to special circumstances, the pores are confined to special 

 areas, we find, of course, that the subdermal cavities undergo a corresponding modification. 

 Thus in Esperella murrayi the subdermal cavities are narrow, horizontally elongated, 

 branching channels (PL XLVIII. figs. 2, 2a) underlying the pore-areas, and in this case 

 they are sufiiciently distinctly marked ofi" from the true inhaleut canals, which lead out of 

 them by narrow openings (PI. XLVIII. figs. 2, 2a). 



In the Clavulina (especially the Suberitidse) we find a good deal more regularity in 

 the form and arrangement of the subdermal cavities, a fact which is undoubtedly due 

 to the presence of a distinct cortex, and to the radiate arrangement of the dermal 

 skeleton. The vertical disposition of the spicules in the cortex prevents extension of 

 the subdermal cavities in a horizontal direction, and they consequently become A'ertically 

 elongated, and we have here an important distinction between the Suberitid^ and the 

 Halichondrina. The pores usually lead in the first instance each into a narrow, slit-like 

 canal, between radiating bundles of spicules (PI. L. figs. 1, 3a) ; these canals then 

 expand, several of them often uniting together, and give rise to a wide space which 

 is the chief part of the subdermal cavity (PI. L. figs. 2, 3a, s.c). Sometimes, as 

 in Latrunculia apiccdis (PI. LI. fig. 1) and Stylocordyla stipitata, var. glohosa (PI. 

 L. fig. 1), this cavity cannot be sharply distinguished from the remainder of the 

 inhalent canal system. At other times, however, it is definitely bounded, its vertical 

 limit coinciding with that of the cortex. Of this latter condition it will sufiice to give 

 two examples, viz., Siiberites caminatus and Tentorium semisuherites. 



In Suberites caminatus the condition of afi'airs is very simple ; each pore leads into a 

 separate, elongated subdermal cavity, which, at first very narrow, gradually increases in 

 diameter, and then, as it reaches the commencement of the choanosome, contracts 

 again more suddenly, being thus somewhat flask-shaped ; occasionally two of these flask- 

 shaped subdermal cavities appear to run into one another. 



In Tento7-ium semisuherites the arrangement is a much more complex one. The 

 subdermal cavities are each divided into three parts corresponding to the three divisions 



