512 GILBERT 0. BOURNE. 



to the species and the place from which the spicule is taken in 

 each species. The colour is deepest along an axial line, and is 

 considerably lighter or altogether absent in the more peri- 

 pheral parts of the spicule. Most of the spicules show traces 

 of a dark core running along their axes. 



The structure of such a spicule may be studied by transverse 

 and longitudinal sections, by crushed and " etched " prepara- 

 tions. A transverse section is shown in fig. 9. Centrally 

 there is a little cluster of dark pores, which comparison with a 

 longitudinal section shows to be the expression of longitudinal 

 canals running along the axis of the spicule. Around this are 

 concentric striations, which become darker and more evident 

 as they approach the periphery. In some of these minute 

 spaces may be detected. The most characteristic feature in 

 the section is the presence of a number of radial cords, some 

 of which are simple, others forked. They originate in the 

 vei'y centre, starting apparently from the central axial canals, 

 and they end peripherally in the warty projections at the 

 surface of the spicule. On examining the section by polarised 

 light one finds, when the Nicols are crossed, that the whole 

 section is bright, but that the radial cords stand out brighter 

 than the remainder when in certain positions. Fixing one's 

 attention on a single cord and the tissue immediately con- 

 tiguous to it, one finds that as the object is rotated the cord 

 becomes alternately brighter and darker without ever being 

 wholly obscured, whilst there is no great change in the sur- 

 rounding substance, though ill-defined dark shadows sweep 

 across it. In certain positions the radial cords, besides being 

 bright, show brilliant iridescent colours, due to interference. 

 1 gather from this that the whole structure, cords and ground 

 substance alike, is crystalline, and that the axes of the 

 crystals composing tlie cords are situated in a different plane 

 from those composing the ground substance. Examination of 

 a longitudinal section (fig. 10) shows the structure of ground 

 substance and cords even better than a transverse section. 

 The axis of the spicule is occupied by a number of very dark 

 lines diverging at very acute angles from the centre. The 



