Sponge-Jauna oj J^orway. 159 



half the circumference, and the flageUated chambers, vvhicli 

 are abundantly present in the mark and sometimes seem to 

 be in connexion with the subdermal cavity by a minute in- 

 current canaliculus. There are otherwise no discernible 

 canals in the mark. The mark is clearly distinguished from 

 the cortical layer of gelatinous connective tissue which repre- 

 sents the non-fibrous layer of the adult cortex. The fibrous 

 layer is at present represented merely by a thin layer of fusi- 

 form fibres in a granular gelatinous matrix, developed from 

 the exterior of the mark and appertaining more to it than to 

 the cortical gelatinous tissue ; it is entirely unprovided with 

 special spicules. The thick spiculated fibrous layer of the 

 adult cortex is thus comparatively late in developing. No 

 pores are yet visible in the skin, which consists of an external 

 wrinkled membrane, witli round nuclei in a layer beneatli it 

 (cuticula and ectoderm ?), and a mesodermic layer of gelati- 

 nous connective tissue, containing pale oval granular cells 

 dispersed through it. The subdermal cavities are lined by a 

 thin membrane with round nuclei imbedded in it with tolerable 

 regularity : this epithelium may be in continuation with the 

 ectoderm somewhere ; but my specimen does not show it. The 

 centre of the mark consists of colourless gelatinous tissue 

 containing irregularly stellate and fusiform cells ; but its outer 

 half is granular, as in the adult sponge, and crowded with 

 flagellated chambers ,* if these are in connexion with a cleavage- 

 cavity, it is curious that there is nothing in my specimen to 

 indicate it. Many Amccha-VikQ cells arc present in the mark 5 

 and in one of them a young acerate spicule is seen almost 

 wholly immersed, as tliough the latter had developed Avithin 

 it (PI. VII. fig. 12) ; and, considering that the cell is almost 

 precisely similar in shape and in the size and character of its 

 nucleus and nucleolus to that in which the Geodia globate 

 develops, this suggestion seems not improbable. With re- 

 gard to the character of its nucleus it also resembles closely 

 the ova of the sponge, but differs in other respects, its outer 

 sarcode being more transparent, less densely finely granular, 

 and staining much more faintly with carmine. Much more 

 close is its resemblance to the Amceba-Y(ke, ceils previously 

 mentioned as associated with some of the large spicules of the 

 adult sponge : and on reexamining these I find that the asso- 

 ciation is much more common than I had before supposed ; it 

 appears \y all not fully-developed spicules of which I could 

 get a gv d view, and not only in Tetilla, but in Geodia 

 Barretti md Isops PJdegrad as well ; moreover, in a great 

 numbe; li cases I could trace from the heap of sarcode which 

 surrounds the nucleus a thin granular film extending towards 



12-^ 



