Sponge-fauna of Norway. 157 



its nucleolus ir^Vu inch in diameter, looking like a globular 

 oil-drop, lies nearer one end of tlie ovum than the other, im- 

 bedded in granular protoplasm, whicli immediately about the 

 nucleus takes a far deeper stain with carmine than elsewhere. 

 At tlie end of the cell, away from the nucleus, yolk-granules 

 (for such I take to be the heap of lai-ge refractive granules 

 represented at g, fig. 5) make their appearance, and increase 

 in quantity with age till the whole cell is crowded with them, 

 except in the immediate vicinity of the nucleus. Some of the 

 yolk-granules appear to present a vesicular form. On the 

 whole, one cannot help being impressed with the similarity 

 of this ovum to that of Hydra. 



The Spermatozoa. — In the three specimens I have examined 

 no trace of these structures was discoverable ; and since the 

 ova occur in every stage of development, the presum])tion is 

 in favour of the sexes being distinct in Tetilla. Large clus- 

 ters of granules occupy a large part of some specimens ; but 

 these are segmentation-spheres of a developing parasite. 



The Cayial-system. — The characters and arrangement of the 

 pores have already been described. They lead directly into 

 the subdermal cavities, whieli extend continuously from one 

 spicular column to another, and communicate between the 

 colunms with each other. In the floor of each subdermal 

 cavity are the inner ends of several ectochones, separated by 

 the usual sphincter from the corresponding endochones ; it is 

 thus clear that each subdermal cavity is equivalent to the 

 outer halves of several ectochones which have become con- 

 fluent, or, vice versuj that those ectochones of a Geodine wliich 

 lie in an area bounded by surrounding spicular fibres are equi- 

 valent to a single subdermal cavity of Tetilla. In SteUetta 

 Normani this is neatly indicated by the subcortical crypts, 

 which have just the same distribution below the fibrous 

 cortex that the subdermal cavities of Tetilla have above it ; 

 they therefore clearly map out the areas which would be 

 occupied by similar subdermal cavities were they present ; and 

 it is interesting therefore to find that they receive from the 

 cortex not one but several endochones between each pair of 

 spicular fibres, as shown in transverse section. The incurrent 

 canals, after entering the mark, soon branch, and continue to 

 branch repeatedly ; but they give oft' at once and all along 

 their course minute short canaliculi, which directly cuter 

 the flagellated chambers. These are joVu inch in diameter, 

 nearly spherical, and very numerously developed. The cliani- 

 bers less abruptly' communicate with the incurrent system 

 by excurrent canaliculi, which are usually longer than the 

 incurrent, the excurrent canals join together into a main 

 Ann. & Mag. N. Hist. Ser. 5. Vol. ix. 12 



