516 GILBERT C. BOURNE. 



whose general direction is parallel to the long axis of the 

 spicule. There can be no doubt that these fibrillae occupy the 

 spaces represented by the axial dark lines or dots in longi- 

 tudinal and transverse sections. From the central bundle 

 fibrillse diverge outwards, and after making several connections 

 with other fibres, whose direction is approximately longitudinal, 

 they run straight towards the surface of the spicule, meeting it 

 at right angles and ending in the warty projections on the 

 surface of the sheath. The whole structure will be best under- 

 stood from an examination of fig. 12. It will be seen that what 

 I have described as the medullary portion of the spicule is 

 occupied by a strand of closely apposed fibres forming a net- 

 work, wliose meshes are greatly extended longitudinally ; that 

 what I have described as the intermediate portion is occupied 

 by a much more open network of fibrillte, and that the cortical 

 portion is traversed by the radial fibrillae, connected only here 

 and there by a longitudinal fibril. The correspondence between 

 the organic and inorganic elements is obvious, and it is clear that 

 the crystalline structure of the whole spicule is dominated by 

 the organic matrix of fibrillse. The needles of calcium car- 

 bonate are arranged parallel to the organic fibrils, and this 

 circumstance explains why the long axes of the crystalline 

 fibres composing the radial cords are set at right angles to those 

 composing the ground substance. Professor Sollas kindly 

 determined the specific gravity of the Spongodes spicules for 

 me, and found it to lie between 2"63 and 2*7. They appear, 

 therefore, to consist of calcite (sp. gr. 2'7), and the lightness 

 of some of the spicules was probably due to their still retaining 

 traces of organic matter in spite of prolonged treatment with 

 Eau de Javelle. 



In the case of the large spicules of Spongodes, Ammot he a, 

 and Siphonogorgia I could find no trace of a cell or cells en- 

 closing the spicule. The last named are deeply embedded in the 

 mesogloea, and no trace of cellular structure could be discovered 

 in connection with them. There can, however, be no doubt 

 that they were originally developed, like the spicules of 

 Alcyonium, Gorgonia, and Clavularia, within cells, and 



