100 



in the star-fishes whose anal aperture is indicated as sub-central. But to ascribe to the 

 presence or absense of this insignificant secretory pore, such a prominent importance as to make 

 it the basis for the formation of larger groups, seems to me to be quite unreasonable. We are 

 thereby forced to separate evidently related forms from each other. As an illustrative in- 

 stance of this we may mention the 2 genera A r chaster and Astropecten, which accor- 

 ding to the earlier division belong to different groups of star-fishes: the former to those 

 furnished with an anal aperture, and the latter to those without it. But these 2 genera are 

 in reality so extraordinarily closely related, that even some species which now are' usually 

 referred to the genus Archaster (for instance A. Andromeda and A. Parelii) have been taken 

 for genuine Astropectens, even up to the most recent times. Only after that the previously 

 overlooked anal aperture (here undoubtedly a mere secretory pore) had been observed in 

 these species, it has been found necessary, in regard to the formerly adopted method of 

 division, to transfer them not only to another genus, but even to an entirely different group 

 of star-fishes. Another distinctive characteristic has indeed been indicated for the genus 

 Archaster, namely the structure of the water-feet; but that it is not possible from this either, 

 to get any trustworthy mark of recognition, has lately been demonstrated by Liitken (1. a). 

 The 2 above-named species of the genus Archaster agree also in this respect completely 

 with the species of the genus Astropecten. 



As has been shewn, the previous division of the star-fishes is scarcely tenable any 

 longer, being based on characteristics of very doubtful systematic value. As principles of 

 division in a new classification of the Asterides, the structure of the ambulacral skeleton, 

 the structure of the cuticular skeleton, and the relation of the other calcareous particles, 

 spines and pedicellaries belonging to the skin, will chiefly have to be considered. Especi- 

 ally the pedicellaries seem in this respect to deserve our attention. They are entirely wan- 

 ting in a great number of star-fishes; while in others (for instance the genus Hippaster) 

 they exhibit a very peculiar structure and arrangement, totally different from what we have 

 become acquainted with in the Brisinga. 



With respect to the structure of the pedicellaries, the Brisinga exhibits most resem- 

 blance to the Asterias and Pedicellas ter, which also agree with the Brisinga in some 

 other points, and may therefore be considered as the nearest relatives of this genus among 

 the star-fishes of the present day. Yet the genus Brisinga exhibits so many peculiarities, 

 that it can scarcely be placed together with those two genera in one and the same family. 

 The Brisinga must therefore in any case be considered as the type for a particular family 

 among the Asterides. With respect to the place of this family in relation to the other 

 families of the Asterides, if we are to follow a definite serial order consistent with the sup- 

 posed phylogenetic dovelopment, it must be placed not at the end of the order of the 

 Asterides, but at the very beginning, as representing the oldest and most primitive of all 

 the Asterides. 



