ASYMMETRY IX THE STARFISH. 135 



has been much speculation as to its morphological and its phylo- 

 genetic significance. Prominent among the speculations that 

 have arisen in connection with this phenomenon are those that 

 concern themselves with the phylogeny of the echinoderms. It is 

 generally held that the ancestral echinoderm was, like the larva, 

 bilaterally symmetrical and that hydrocceles and their derivatives 

 occurred on both sides. The objection to this view is that such 

 an organism would be biaxiate, with two mouths and oral surfaces, 

 two water-vascular systems and nerve rings. In view of this 

 serious objection, it would seem best to look upon the develop- 

 ment of the new axis as an evolutionary process resulting from 

 the establishment of a pronounced asymmetry, and that it was 

 only in connection with this newly established asymmetry that 

 the development of such structures as the new mouth, the nerve 

 ring, the Aristotle's lantern in echinoids, and similar structures, 

 takes place. This view does not preclude the possibility of an 

 originally bilateral ancestor, but merely denies to such an ancestor 

 the bilateral possession of such specialized structures as the 

 hydroccele and its derivatives that arise only in connection with 

 the apical region of an entirely new axis. 



When, however, larvae are found with hydrocceles and kindred 

 structures on both sides, evidence seems to be afforded for the 

 theory that the ancestral echinoderm had two sets of hydroccele 

 structures bilaterally arranged. If such anomalous larvae are to 

 be considered as atavistic, how can we explain the occurrence of 

 larvae in which the hydroccele and its appurtenances appear 

 exclusively on the right side? Is the right-handed or reversed 

 larva also an atavistic reminiscence of an ancestor that had right- 

 handed asymmetry? Even the most obdurate phylogenist would 

 hesitate to go so far ; yet the occurrence of reversed asymmetry 

 is commoner than is bilaterality, and both can be induced by the 

 same experimental means. Moreover, the two conditions grade 

 into each other; for there are many larvae in which hydropores 

 and such structures occur on both sides but in which one or the 

 other side shows these structures in all states of partial rudi- 

 mentation. 



How then can we explain the occurrence both in nature and in 

 experiments of reversed and bilateral larvae? The previous parts 



