Sponge-fauna of Norway. 143 



(v) The subcortical layer is similar to that of Geodia 

 Bnrretti. 



2. The Mark. — Tliis is chiefly composed of a granular 

 connective tissue like that of other Geocline sponges, but 

 partly also of vesicular connective tissue ; in places groups 

 of granular protoplasmic cells, containing a number of 

 highly refractive globules resembling oil-drops, are met 

 with. Tlie ciliated chambers measure about O'OOl inch in 

 diameter. 



3. The Canal-system. — (i) Incurrent canals. Commencing 

 with an examination of the surface of the sponge, we find that 

 pores are not universally distributed over it, some parts being 

 quite destitute of them ; and in these places transverse sec- 

 tions of the crust prove the corresponding absence of chones. 

 In the poriferous surface we can frequently distinguish a 

 number of roundish or polygonal opaque white areas, sepa- 

 rated by slightly more translucent interspaces ; in these areas 

 are set the pores, a group of six to ten pores in each, though 

 sometimes only one or two are visible, or, it may be, even 

 none. If the tissue bearing the pores be removed from the 

 sponge and examined in glycerine under the microscope 

 (PI. VI. fig. 4), it will be found to consist of a layer of dermal 

 vacuolated tissue, covered by the epidermal and bacillar 

 layers; between adjacent groups of pores, and serving to 

 define them from each other, a row of globate spicules re- 

 placing the dermal tissue is seen in addition. The dermal 

 tissue forms a thick framework between the pores, but thins 

 out towards their margins, leaving these to be constituted by 

 the epidermal and bacillar layers alone. It is quite clear 

 that these pores, although doubtless capable of opening and 

 closing by iris-like movements of the clear marginal mem- 

 brane, are not vague and transitory, as has been asserted, but, 

 on the contrary, persistent and well defined. In my speci- 

 mens they are usually elliptical in form. With regard to 

 their size, concerning which much confusion exists in pub- 

 lished writings, I find that tlie diameter averages about 0'075 

 inch ; sometimes it becomes as much as 0*12 inch, or as little 

 as 0"002 inch. The majority are clearly visible to the naked 

 eye. On examining the surface from wliich the poriferous 

 layer has been removed, it will be found that a chonal cavity 

 lies beneath each cluster of pores ; the opaque white areas 

 in which they are set are thus in fact chonal roofs, the opacity 

 and whiteness being due to the absence of the globate layer 

 beneath them. 



The chones, of which we have next to speak, are closely 

 similar to those of Geodia Barretti] they were first described 



11* 



