REPORT ON THE ASTEROIDEA. xxvii 



and at the adult stage have become more or less insignificant. It may hence be stated 

 briefly that the members of the more highly developed group (the Cryptozonia) pass in 

 the course of their development through a stage which represents the characters of the 

 adult condition of the more primitive group (the Phanerozonia). 



Of the development of intermediate plates between the supero-marginal and infero- 

 marginal series, and of the relative posture of the plates in these two series in the Crypto- 

 zonate Asterids, it is unnecessary here to remark. 



From the foregoing considerations it will be seen that three of the most important class 

 characters of the Asteroidea exhibit two modes of presentment, and that by means of any 

 one of the three the Euasteroidea may be divided into two groups, one of which in each 

 case is apparently older than the other. 



Now the divisions thus made are essentially equivalent in each of the three categories ; 

 that is to say, the older group as determined by one of the characters corresponds to the 

 older group determined by the other two, and similarly with the newer and more highlv 

 developed group. The corresponding divisions in each of the three categories are not, 

 however, exactly coequal when regarded independently, nor is this to be expected ; but 

 the slight overlap or extension of one or other of the characters in the case of transitional 

 groups is comparatively insignificant : in fact, the result is a "dove-tailing" or interlocking 

 of the groups, which is perfectly natural and in accordance with what is found to occur 

 in other branches of the animal kingdom. We have thus a natural morphological division 

 of the sub-class into two sections or groups on the basis of three independent and im- 

 portant structural factors, one group being more archaic or more primitive than the other. 

 These groups I regard as true natural orders in the strictest sense of the term. To the 

 first or older group I have given the name Phanerozonia, to the second and more highly 

 developed group the name of Cryptozonia. 



I have selected the marginal plates for the name character of the orders in 

 consequence of their importance in the Asterid skeleton as well as for the compara- 

 tive ease with which their character may be observed in the superficial aspect of a 

 starfish. 



1 shall now proceed to define the families comprised in the two orders, in accordance 

 with my views as to their morphological relations, and shall cite under each the consti- 

 tuent genera. I have been led on structural grounds to establish subfamilies where the 

 divergence of form indicated by the included genera demanded in my opinion this 

 expression of their affinities. In many cases this presentment of the constitution of a 

 group will be found to give a solidarity to the family and an intelligent view of its com- 

 position, as well as to indicate the relationships of genera which otherwise would appear 

 isolated and unnaturally placed. As an evolutionist, however, I regard the ideas of 

 species, genera, families, and other divisions, as merely abstract particulars which con- 

 stitute a synopsis or working key to our views of the affinities of organic forms — the 



