INTERRELATIONSHIPS 6l 



Second, by the evidence furnished by the new researches on the larva of a 

 comatulid of the Actinometra type described in a later chapter, showing that 

 the radianal actually begins in the axil between the two basals before the radial 

 is formed, and is thus at the base of the yet undeveloped right posterior ray; 

 from this position, with the development of the radial, the radianal works its 

 way from below the radial and passes upward to the left of it. 



Returning now to the form assumed in the hypothetical figure (PI. A, 

 fig. 7), — the lines of modification from such a form on the dorsal side would be: 



a. In the radianal, by migration upwards. 



b. In the anal system, by ( 1 ) extension upward in vertical series not con- 

 nected with brachials, except by perisome; (2) increase upward and laterally 

 in sutural connection with the brachials; (3) elimination of the anal plate by 

 absorption, atrophy, or migration or shifting. 



c. In the brachial system, by ( 1 ) increase in the number of primibrachs ; 

 (2) variation in the mode of branching of the higher brachials or arms. 



d. In the interbrachial system, by growth and multiplication of supple- 

 mentary plates between the rays. 



On the ventral side we might expect modification in the same way that is 

 seen in Antedon, viz., a growth of the perisome between the radials and orals, 

 gradually separating the latter plates from the radials and carrying them 

 inward toward the center. This would give the condition found in Onych- 

 ocrinns and Taxocrinus; but the cases in which these structures are preserved 

 in the fossils are too few to enable us to trace any of the successive changes. 



a. The first of the above modifications of the dorsal side has already been 

 discussed; and we are able to trace in the phylogeny of the Flexibilia succes- 

 sive stages in the migration of the radianal, from that shown by the hypothet- 

 ical figure to that of elimination from the calyx, which form the basis of 

 important generic groups. These stages are analogous to phases in the de- 

 velopment of a corresponding plate in the ontogeny of a living crinoid, as above 

 stated. 



b. The generic distinctions based upon the other modifications of the 

 posterior, or anal, area coming under the second category, are really very 

 striking — more so than can be well expressed in terms of brief analysis. Aside 

 from the matter of the radianal, there are two plans of structure of the anal 

 area which run side by side from the Silurian to the Carboniferous. They 

 start with the primitive anal plate of our supposed ancestral form, and diverge 

 upon the two lines 1 and 2 indicated above. 



I. The first of these plans is that which represents a solid support or back- 

 ing of an anal tube. It is marked by a vertical row of strong, rounded plates, 

 originating on the posterior basal, rising with a very gradual taper to a con- 

 siderable height between the rays, and having the appearance of a small, 



