174 Dr. H. Eaiiff o^e the Oenus Hindia, Dune. 



canals are indicated as L, II., and III., and their similarly 



placed surfaces as a, h, c, d, e,f; so that II. a coincides with 

 I. d, lll.h with I. e, II. /with III. c, and so on. 



The central points of the spicules are then situated in the 

 angles of the six-sided columns, tliose of the neighbouring 

 radial series alternating as already stated, the rays 1, 2, 3 in 

 the three surfaces meeting in one angle respectively; for 

 example, I. c — I. c?(=II. a) — II. b (ray 2 not visible, because 

 it is situated in the surface I. d, which is turned away from 

 the observer), or I.a—l.b — IV./ (IV. indicating another 

 canal united to I. in the surface I. h), or I. b — I. c — IV. d, and 

 so on, the longitudinal direction of the fourth aborted arm 

 (omitted throughout in fig. 2) always coinciding with the 

 an^le of the prism. The arms no. 4 of all the spicules placed 

 upon the same angle w^ould therefore, if they were developed, 

 unite the central points of the spicules, as it were materializing 

 the angle and further dividing the rnesh-spaces in a radial 

 direction. I have indeed sometimes thought that I observed 



