146 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



tho liidials as a base, and they have been heretofore universallj' so considered. In 

 this case the existence of two plates between the bases of the two posterior in many 

 types would be a fact of considerable morphok)gical significance as designatmg a 

 fundamentally differentiated area; but according to the latter interpretation the 

 five interradial areas, mcluding half of the ambulacral system on either side of each, 

 are the true units of pcnitamerous spnmetry, and therefore the existence of addi- 

 tional plates m one of th(! interradial areas merely indicates that the two borders 

 of this area have for some reason or other become somewhat more separated than 

 those of the other four, necessitating the development of protective })latcs to cover 

 the exposed perisome, the occurrence of such plates having a fundamental mor- 

 phological signifiQance no greater than that of polydactylism of a single limb among 

 the vertebrates or arthropods. 



It must be constantly borne in mind that there is absolutely no direct corre- 

 lation bet'^'een the primarily skeleton forming dorsal surface of a crinoid and the 

 pruuarily perisomic ventral surface and the (secondarily) superficial ventral internal 

 organs. 



The skeleton of the dorsal surface and the dorsal nervous system are governed 

 in their arrangement entirely by the heredity and by the ancestral meristic division, 

 the somatic divisions, here consisting each of an interradial area with half of the 

 adjacent radial areas or ambulacral areas as borders, constituting the five half meta- 

 meres of which tho crinoid is composed. A secondary rearrangement both of the 

 calcareous structures and of the nerves has taken place which to a large extent 

 masks this original arrangement, especially in the elongate body processes, but it 

 may always be detected on close examination. 



The prolongation of the closely apposed marginal plates of the five origmal 

 metameric divisions into arms offered an oi)])ortunit3^ for the extension of the ring 

 systems about the CEsophagus in five long radial lines, of which advantage imme- 

 diately was taken ; or, to express it in another way, the arms in their elongation have 

 drawn out into long processes lying upon their ventral surface the radial tliverticula 

 from the radial circumoral systems with which they are, on account of their phylo- 

 genetical and ontogeiietical origin, most intimately and indissolubly connected. 



Thus there is a marked secondary correlation of very recent origin within the 

 class between the dorsal and ventral systems which is the result of economic possi- 

 bilities afforded by the intersomatic (radial) extensions of the dorsal system to the 

 ventral systems. 



In the primitive phyllopods the body consists of a large but varWng number 

 of segments which are remarkably uniform in structure, but in the remaining 

 groups the segments become locahzed in definite and strongly marked body divi- 

 sions;. in those tho most ustial numbcn- of significant somatic divisions included 

 within the thorax is five (well illustrated in th(> Decapoda) and this fact is seen to 

 be of no little importance when we realize that the echinoderms are essentially one- 

 half of a fivc-segmonted crustacean thorax from which the head and the abdomen 

 have disappeared by atrophy concurrently with the missing side. In this connec- 

 tion the greatly overdeveloped thorax of the majority of the crustaceans, and the 

 entire degeneration of the head of others, should be noted. 



