12 VERRILL 



MULTIPLICITY OF RAYS ; VARIABILITY. 



It is a singular fact, for which no satisfactory reason has yet been 

 given, that nearly all Echinoderms, whether living or fossil, are nor- 

 mally five-rayed. This rule prevailed quite as constantly in the 

 remote paleozoic ages as in the present period. 



The number of rays was apparently well fixed in the unknown 

 primordial ancestors of the earliest fossil Echinoderms of most 

 classes, for all classes of Echinoderms, except, perhaps, the Holo- 

 thurians, appeared in the early geologic ages, with most of their 

 more important features much as we find them now. 



Among Paleozoic Echinoderms, as among modern ones, certain 

 starfishes departed from the five-rayed type by acquiring additional 

 rays. As in some modern genera, this was apparently done in post- 

 larval life by the interpolation, or budding in, of new rays between 

 the older ones, in some of those species having numerous rays. 



At present, this is known to occur only in Pycnopodia of the 

 Northwest coast ; in Heliaster, represented by several species on the 

 tropical and subtropical Pacific coasts, from lower California to 

 Chile ; and in Labidiaster, from Patagonia. It is probable, also, that 

 it occurs in Rathbunaster Fisher, a deep-water Californian species, 

 allied to Pycnopodia. 



Of Heliaster, seven species are known. When adult, the number 

 of rays exceeds twenty and in some species is as high as forty to 

 forty-four. 



In all its species new rays are gradually interpolated, rather 

 irregularly, between the older ones, all around the circumference, but 

 not without some order. 1 



The same is true of Labidiaster, which has numerous long rays 

 when adult. 



In Pycnopodia, the number of rays when adult is twenty to twenty- 

 four. It normally starts, when very small, with six equal rays. A 

 new pair of rays then appears, one on each side of one of the primary 

 rays ; then another pair appears just back of these, and so on in suc- 

 cessive pairs. 2 This would always produce an even number of rays. 

 But variations from this bilateral regularity often appear, producing 

 odd numbers. (See pis. LXXIII and LXXIV.) 



' See H. L. Clark, Starfishes of the Genus Heliaster, Bull. Mus. Comp. Zool., 

 vol. LI, pp. 25-76, pis. i-vni, 1907. In this paper all the species are fully de- 

 scribed and their modes of adding new rays are explained. 



" See plates. See Professor W. E. Ritter and G. R. Crocker ( Proc. Wash- 

 ington Acad. Sci., vol. n, pp. 247-274, 1900), who give a good account of the 

 process. 



