REVISION OF PALEOZOIC STELLEKOIDEA. 39 



On the other hand, the increase of rays is more probably due " to 

 the advantage gained in holding their food securely, and in opening 

 bivalves, than for holding to the rocks, though both go together." 



"The ability to cling tenaciously to rocks may be perfected in 

 other ways, involving equally an increased number of sucker-feet. 

 This is often attained by lengthening the rays, * * * by crowding 

 the suckers into more than four rows, * * * and by increasing 

 the size and strength of the suckers." 



DEVELOPMENT or THE PARTS. 



Anus. In all living starfishes the anus apparently lies in the next 

 interradius to the right of the madreporite, or in the direction as the 

 clock goes, but in no Paleozoic asterid is there positive evidence of 

 an anal opening. However, it is extremely rare in these forms to find 

 the disk so well preserved as to make it possible to locate so small 

 an orifice. In certain specimens of Hudsonaster, immediately adj acent 

 to the centro-dorsal plate, there is an open space that either is the 

 anal opening or where a plate has dropped out in fossilization. 

 Schondorf has also failed to find this structure in the German Paleo- 

 zoic asterids, but has indicated its probable position in Spaniaster 

 (see fig. 8) and Calliasterella. 



Madreporite. In more than half of the Paleozoic asterids the 

 madreporic plate has not been seen. In many cases this is due to the 

 infallen condition of the disk skeleton and the jumbled state of the 

 ossicles; in others it is due to the fact that it is so small as not to be 

 readily distinguished among the mass of other small plates. Finally, 

 a great number of asterids are known only from the actinal side. 



The madreporite in size varies from minute to very large and 

 ponderous. It may be round, oval, or many-sided, concave, and 

 depressed beneath the disk pieces, or protruding cone-shaped (Petras- 

 ter) . As a rule, it is marked by striations that are coarse or very fine, 

 straight or wavy lined, or it may be crenulostriate or with granular 

 surface. In Petraster the surface canals enter into two spirals that 

 carry the water into the stone canal. In living forms with more 

 than six rays there may be two or more madreporites, but in no 

 Paleozoic species has more than one been seen. 



In all of the Paleozoic phanerozonians the madreporite appears 

 to be always dorsal in position. In the most primitive genus, Hud- 

 sonaster, the plate may be large and conspicuous or at times can not 

 be made out even when the disk is fairly well preserved. In no 

 Mesopalseaster has this sieve plate been seen, although good speci- 

 mens are at hand, a condition probably due to its small size among 

 the small skeletal pieces of this genus. In Spaniaster it is very small. 

 In Promopalseaster the madreporite is usually conspicuous and large, 



