514 GILBERT 0. BOURNE. 



isolated behaves as a single crystal under the polariscope. 

 The fibrous structure, however, is better shown in etched 

 specimens. 



A spicule of Spongodes, which has been carefully treated 

 with very dilute acid, breaks up into bundles of exceedingly 

 fine acicular crystals, arranged with their long axes nearly 

 parallel to the long axis of the spicule ; they are not quite 

 parallel, but are fitted together and interwoven so that any 

 one bundle of fibres gives brilliant interference colours when 

 viewed through the polariscope. The radial cords are not 

 easily studied in etched specimens, but it can be seen that the 

 longitudinal feltwork of crystalline fibres is interwoven with 

 the radial cords much as a fabric would be which had very few 

 and coarse strands in the warp, and very numerous and fine 

 strands in the woof. 



When a longitudinal section of a spicule is examined by 

 polarised light with the Nicols crossed, the radial cords are 

 seen, in certain positions, to stand out brightly; whilst the 

 ground substance is nearly dark, though lit up here and there 

 by iridescent tints. As one rotates the section the radial 

 cords become dark, whilst the ground substance, still retaining 

 its iridescence, becomes lighter, and these alternations take 

 place regularly for every 90° through which the section is 

 rotated. It has been shown that the ground substance is 

 composed of nearly parallel longitudinal crystalline fibres, and 

 it may be inferred that the radial cords are composed of similar 

 crystalline fibres, whose axes are at right angles to the fibres 

 of the ground substance. 



A single crystalline fibre measures someO'03 mm. in length, 

 and has the optical characters of a single crystal, being extin- 

 guished in four positions at right angles to one another when 

 rotated between the crossed Nicols. 



As none of my sections were thin enough to obtain a section 

 of a radial cord with two plane surfaces, and as the cords are 

 neither straight nor of uniform diameter (see figs. 9 and 10), 

 but swollen in places so as to be almost moniliform, they ex- 

 hibited interference phenomena under polarised light in every 



