THE FOSSIL CRIISrOID GENUS DOLATOCRINUS AND ITS ALLIES. 25 



The number of visible openings depends upon the extent to which 

 brachials are incorporated, as has been shown in regard to the two 

 species of Comanthocrinus. If the arm becomes free on the first 

 secundibrach, there may be no pores through the calyx wall, or only 

 a single one at one or both sides of the arm-base ; if higher secundi- 

 brachs are incorporated, the number correspondingly increases. In 

 D. grandis (pi. 7, figs. 2, 5) the biserial arm is incorporated for a 

 distance of several pairs of brachials following the third secundibrach; 

 a fixed pinnule leads from IIBr^, on the outer side of the dichotom. 

 followed by one on IIBrj, inner side, next by one on IIBrg, outer 

 side, and then by one for each margin of the biserial pair, each of 

 which leads to an opening. Therefore when one pair of brachial 

 ossicles succeeding IIBrg are incorporated, there will be 6 pores 

 between the rays and 4 between their divisions; and with further 

 incorporation of arm brachials the number of openings will be 

 increased to 8, or even 12. In D. hulbaceus only one small pore is 

 present for each arm; in stelUfer and am plus there are always 2 for 

 each arm, and from 2 to 4 for each interradius. 



Numher of arms. — With the foregoing category of available char- 

 acters to choose from, it ought to be practicable to find correlations 

 of sufficient constancy to define the species without resorting to any- 

 thing so purely arbitrary as the number and grouping of arms, 

 standing alone, in forms wJiere this character is clearly subject to 

 great individual variation. I do not wish to be misunderstood on 

 this point, and it must not be supposed that I am denying the impor- 

 tance of the number of arms as a general proposition. On the 

 contrary, it is a character of great value and of wide application, 

 representing as it does progressive stages in the ontogeny of the 

 crinoids which often become fixed, and furnish decisive criteria in 

 the separation of groups. The entire Larviformia division of the 

 Inadunates is characterized by having the arms in the primitive 

 stage, without any axillary plate in the ray, and therefore five in 

 number. The change in other groups from 10 to 20 arms, due to a 

 second bifurcation in all the rays, indicates a certain progress in calyx 

 development by increase in ray capacity that probably correlates 

 with other facts. Where the change is of less degree — such, for 

 example, that only one of the two arms in each ray bifurcates— but 

 according to a definite plan, it may be a good character. For 

 instance, the 15 arms, 3 to each ray, of D. triangulatus of the Michigan 

 area, constant as I have it in 23 specimens, and correlated with a 

 definite form and proportion of calyx and style of ornament, is a 

 good character to help distinguish it from somewhat similar forms 

 having a constant number of 2 or 4 arms to a ray. The 10 arms of 

 D. grandis is a perfectly reliable character as against the 20 arms 

 of D. spinosus, or the 30 arms of D. exstans, or the 40 of D. multi 

 hracJiiatus. But if a specimen of D. spinosus happened to have 



