36 



I'ANAMIC DEEP SEA ECHINI. 



primary and other tubercles across adjoining sutures or over adjoining 

 smaller tubercles and miliaries. 



It is probable that in an earlier stage the anal system was only covered 

 by five plates. The specimen of 1.9 mm.. Fig. 48, had seven triangular 

 plates, five of equal size, four of which carried miliaries. and two slightly 

 smaller (PI. 21. fig. ..'). In an older specimen •"> mm. in diameter, Fig. 49, 

 the anal system is hexagonal (PI. 21, fig. ft), with an outer row of seven 

 large plates which have lost their triangular shape and have become mainly 

 hexagonal, with an inner row of seven elongate pointed small plates in the 

 angles of the primary anal plates and three smaller, narrow plates. In a 



V ' i 



1.9 mm. 5 mm. 



Fig. 48. Salenia varisfina. Fig. 49. 



5 mm. 



Fig. 50 Salenia miliaris. 



specimen of S. miliaris, also measuring 5 mm. in diameter, Fig. 50, the anal 

 plates are more regularly arranged (PI. 16, fig. ..'), there being five large 

 primary anal plates, an inner row of five smaller ones intercalated between 

 the primary plates, and a third set of still smaller, narrow, elongate plates 

 in the actinal angles of the second set of plates. So that it appears 

 that the anal system of Salenia, as of Cidaris, was originally covered by 

 five plates. 



The buccal ambulacral plates (PI. 21, fig. 1) occupy nearly the whole 

 of the actinal system. They touch laterally, and are separated from the 

 actinostome by a belt of two or three rows of calcareous plates, slightly 

 imbricating, irregularly arranged (PI. 21, figs. 1, 4). The plates of the 

 actinal system are made up of an open network of cells. The ambulacral 

 pores are not as yet equally developed (PI. 21, fig<. 7, 4), five of the 

 suckers being more prominent than the others, the pores being scarcely 

 visible on the other buccal plates. 



The Salenia Goesiana figured by Loven 1 has unfortunately lost its actinal 

 and anal plates, as well as radioles, so that we can only compare the test to 



1 Loven, Echino'dees, p. 27, PI. XIX. 



