654 THE SPINES. 



/. 8). In the fossil genus Acrosalenia (PL. XXXV. f. e) we find the nearest 

 approach in the structure of the shaft to Arbacia, only without the charac- 

 teristic edge. It recalls also the structure of the recent Phymosoma (PL 

 XXXVI. f. 11), but has not, like it, the well-marked circular rings, denoting 

 the different stages of growth of the spines, which is a structural feature 

 characteristic of all the other Echini — whether Desmosticha, Clypeastroids, 

 or Petalosticha (with the exceptions just enumerated) — which 1 have had 

 occasion to examine, and the sections made extend to the principal genera 

 of these suborders. In the Diadematidae we have a combination of two 

 modes of growth, of the concentric and of the cellular. Near the base of 

 the spine, which may bo hollow [PL XXXV. f. i<) in the primary interam- 

 bulacral ones, or solid (PL XXXV. f. 15) in the small ambulacral spines, 

 the section is made up of a couple of rings ; the inner ring consists of spokes 

 of small cells, irregularly arranged between the large outer solid wedges ; 

 this is then succeeded by a second ring of irregularly arranged cells, edged 

 by a row of small solid wedges (PI. XXXV. f. 1,:) ; higher upon the shaft 

 of the spine the solid wedges have become connected, and are separated 

 by the spokes of minute cells (PI. XXXV. f. 11). The wedges form the 

 verticillations so characteristic of the Diadematidae, and, in fact, they con- 

 stitute the ornamentation of all the spines, whether they form an indepen- 

 dent sheath, as in the Cidaridae, growing in size at the same time and inde- 

 pendently with the central part of the shaft, or. as in the other Echini, form 

 a part of the inner substance of the shaft, and increase at the same time with 

 the cellular interlamellar substance. 



In the Echinometradae we find the concentric rings most distinctly marked. 

 Each ring terminates by a row of larger cells, indicating clearly the growth 

 to have taken place at definite intervals, until the spines have reached a 

 maximum size ; but even then, as has been well shown by Carpenter, they do 

 not cease to grow, as huge broken spines of Ileterocentrotus are very soon 

 repaired, and attain frequently a larger size than they would under ordinary 

 circumstances. 



The great size of the wedges and their intermittent growth are especially 

 marked in sections of Echinometra (PL XXXVI. f. 1), of Stomopneustcs 

 (PL XXXVI. f. 2), and of several species of Strongylocentrotus (PL XXXVI. 

 f. 4, 9). In Heterocentrotus (PL XXX VI. f. 8) and Podophora the wedges 

 are small, connected ; hence the smooth or finely granular surface presented 

 by the outer surface of all the spines of these genera. The Echinidae proper 



