Sponge-fauna of Norway. 153 



the same kind of oval annular wall before described. Hamate 

 spicules are also found lying beneath the outer membrane. 

 The interpretation of these observations is most difficult. If the 

 outermost membrane be the ectoderm, why do not at least 

 nuclei appear in it ? since they are plainly seen in the sub- 

 dermal ectoderm, and these have the same appearance as the 

 nuclei which here occur below the outer membrane. If it is 

 not ectoderm, but a structureless cuticula, where is the ecto- 

 derm? for it appears very improbable that a heterogeneous 

 layer of oval cells, spicule-cells, and bare nuclei can represent 

 this layer ; by no means impossible, however, since, in spite 

 of the beautiful demonstrations we have had from the masterly 

 hand of F. E. Schulze, I do not yet feel quite convinced of 

 the uniformity of this structure over all parts of the sponge, 

 or that it presents the same constancy as in the other Metazoa. 

 But it is reckless to speculate in the absence of any of the 

 evidence possible, and the silver treatment may eventually, 

 as I almost expect, bring out of the apparently structureless 

 cuticula the polygonal outlines of epithelial cells ; but, till that 

 welcome sight appears, one must be content to take the facts 

 as one finds them ; and so provisionally the outermost layer 

 is for me a cuticula, and the mixed cellular layer beneath a 

 heterogeneous ectoderm. The cuticula and ectoderm together 

 cover the exterior of the sponge, except in some cases near 

 the small oval pore-areas (PI. VII. fig. 15), which are situated 

 in the meshes of the polygonal dermal network ; in these it 

 often happens that the cuticula is absent, or has thinned away 

 beyond one's power of observation ; for, although present on 

 the surrounding skin, it is not possible to see what becomes 

 of it as it is traced into the pore-area. The membrane in 

 which the pores in these instances are set appears, when 

 looked down upon from above, to consist of a structureless 

 film which does not stain with carmine, and in which are 

 imbedded granules, naked nuclei, similar nuclei surrounded 

 by hamate spicules which often project out of the membrane, 

 and flat oval-walled cells with similar nuclei ; thus it has 

 the same appearance as the ectoderm of the subdermal cavities. 

 In addition, fine delicate filaments may be seen, which fre- 

 quently run parallel in groups, crossing one another in the 

 space between two pores, and then diverging so as to touch 

 tangentially the edge of the pores. In the dermal network 

 which forms the floor of the oscular tube and the roof of the 

 subdermal cavity below, Ave can perceive a similar structure 

 (PI. VII. fig. 8) ; but transverse sections show here a super- 

 ficial ectodermal layer with a thin mesodermic layer between, 

 where the network is thickest ; in the very thin layer which 



