306 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



processes; for in each case there has been simply an elongation of the cirrus, the 

 produced tip remaining of the same typo as the basal portion in the first instance, 

 but acquiring flexibihty, and hence inducing a finer division of the primitive 

 homogeneous calcareous investment, in the second. 



Both these processes may often be traced in a single specimen; for the short 

 cirri at the dorsal pole of the centrodorsal (fig. 310, p. 269) are really the persistent 

 cirri of the young which were formed at the time when the ventral rim of the 

 centrodorsal was only just anterior to the proximal (upper) border of their sockets, 

 and the succeeding cirri were hkewise formed as the centrodorsal gradually 

 increased in size through additions to its ventral rim, each row of cirri representing 

 the stage at which the centrodorsal was only the equivalent in size of that portion 

 of the adult centrodorsal l)etween the upper margin of that row and the dorsal pole. 



By a study of the succession of the cirri in good specimens of Leptometra, 

 TTiysanometra and Nanometra (fig. 310, p. 269) it is at once evident that in all cases 

 the cirri were at first of the type seen, in a slightly modified form, in Antedon medi- 

 terranea (figs. 105, p. 169, and 313, p. 271), but have become gradually modified 

 along the lines described until the adult type has been attained. 



The sequence of the added segments in these forms is the same as that described 

 in the Thalassometridie (p. 290), but with the difference that in the Thalassome- 

 tridae, as in most of the OUgophreata, there was a crystahzation of the type of cirrus 

 at or near the stage seen in the Charitometridse (figs. 99, p, 160, and 100, p. 162) 

 and in Tropidmetra (fig. 356, p. 293), and the change from the short stout and 

 smooth tj-pe to tlie long, more slender, and spiny type was effected by a cumula- 

 tive phylogenetic force, restrained for a long time by the inertia of long-estabhshed 

 habit of form, which finally burst its bonds and all at once gave rise to the per- 

 fected cirri, such as are seen in the Thalassometridae (figs. 93, p. 153, 94, p. 155, 

 95, p. 157, and 96, 97, p. 159). The Macrophreata were much more plastic, and 

 had no primitive fixed cirrus type, so that cirrus development has progressed 

 evenl}' without any sudden eruption of long pent up phylogenetic force, and each 

 stage shows merely a uniform and slight advance over the preceding. 



There is no correlation whatever observable between the type of cirrus and the 

 character of the centrodorsal except in such secondary ways as where an increase 

 in the size of the cirri is accompanied by a corresponding increase in the size of 

 the centrodorsal, but without any other change in its general form. 



Long cirri wdth comparatively long segments proximally and very short seg- 

 ments distally are found irregularly placed in from one to three rows on a hemi- 

 spherical or thick discoidal centrodorsal sho'sving no radial resorption in: 



Comantlius (part). Oxymetra. 



Zygomeira (part). Dichrometra (part). 



Amphimetra (part). Cenometra. 



Himerometra (part). Colohametra. 



Ileterometra (part). Cyllometra. 



Pontiometra. Decametra. 



