364 W. J. SOLLAS ON THE STEFCTUEE AND 



one over each of the four inclined edges of the pyramid. The base- 

 of the pyramid must be regarded as rectangular, a little longer than 

 broad ; for even in the uncrushed state the rays of the spicules dO' 

 not form right angles with each other, but include one pair of oppo- 

 site angles of less, and another of more than 90°. Probably, how- 

 ever, this departure of the rays from strict rect angularity is due to 

 the pressure which brought about the cleavage of the slate in which 

 the spicules occur ; for it is noticeable that the spicules of any single 

 specimen are all similarly orientated; i.e. the larger angles all look 

 one way, and the smaller another, as if some general distorting force 

 had acted upon all the spicules at once, closing up the rays in one 

 direction and opening them out in another. 



Four rays are all that each spicule can be shown to possess ; but 

 a fifth might naturally be expected to be present, having its origin 

 from the under surface of the spicular centre, and a position coin- 

 ciding with the morphological axis of our imaginary pyramid. 

 "Whether such a ray actually exists or not, I have been unable to 

 determine, though certain obscure indications lead me to believe that 

 it does. On the other hand, it is quite clear that no ray was given 

 off in the opposite direction, i. e. from the upperside of the middle 

 of the spicule upwards ; for this side is perfectly preserved and clearly 

 visible, and yet affords no trace of the existence of such of a ray ; it 

 has evidently been always as smooth and devoid of ornament as it is 

 now. 



The spicular rays are quite circular in section and almost com- 

 pletely cylindrical in form, scarcely tapering at all, so far as they 

 can be traced — except in the case of the smaller spicules, the rays of 

 which are evidently acute cones ; none of the spicules, however, shows 

 the ultimate termination of a single ray ; for the ends of the rays are 

 either broken off, imbedded in the slate, or hidden by passing beneath, 

 the arms of adjacent spicules. In size the spicules are of four dif- 

 ferent grades : in the largest forms the spicular rays are -i-S- of an inch 

 long, so that the whole spicule measures as much as |-^- of an inch 

 between the ends of two opposite rays ; in the second-sized spicule 

 the rays are -£-^y■ of an inch long, in the third -£jy, and in the fourth 

 size -g\ of an inch long ; thus each spicule is twice the size of the 

 one next succeeding it in the scale. 



Arrangement of the Spicules. — The largest or primary spicules 

 form a framework, which is filled in by the smaller forms. The 

 secondary spicules are placed so that the centre of each coincides 

 with the centre of the square which would be formed by completing 

 the rectangle included between the adjacent rays of the primary 

 spicule. The four rays of the secondary spicule proceed from their 

 origin directly towards and at right angles to the sides of the imagi- 

 nary square, one to each side ; those which approach the rays of the 

 primary spicule descend beneath them and become lost to observa- 

 tion. The tertiary spicules are similarly arranged ; but in their case 

 the squares within which they lie are chiefly real and not imaginary, 

 some of the squares being formed by two rays of the primary and! 

 two of the secondary spicules, others having two sides formed by 



