THE RECAPITULATION THEORY 353 



precociously developed. The latter explanation is 

 the one usually adopted ; but before the question 

 can be finally decided more accurate observations 

 than we at present possess are needed concerning 

 the stages intermediate between the egg and the 

 Nauplius. The absence of a heart in the Nauplius 

 may reasonably be associated with the small size 

 of the larva. 



Concerning the larval forms of vertebrates, it is 

 only in Amphioxus and the Ascidians that the 

 earliest larval stages are free-living, independent 

 animals. In both groups the most characteristic 

 larval stage is that in which a notochord is present, 

 and a neural tube, open in front, and communicating 

 behind through a neurenteric canal with the digestive 

 cavity, which has no other opening to the exterior. 

 This is a very early stage, both in Amphioxus and 

 Ascidians ; but so far as we know, it cannot be 

 compared with any invertebrate larva. It is 

 customary, in discussions on the affinities of verte- 

 brates, to absolutely ignore the vertebrate larval 

 forms, and to assume that their peculiarities are 

 due to precocious development of vertebrate cha- 

 racteristics. It may turn out that this view of the 

 matter is correct ; but it has certainly not yet 

 been proved to be so, and the development of both 

 Amphioxus and Ascidians is so direct and straight- 

 forward that evidence of some kind may reasonably 

 be required before accepting the doctrine that this 

 development is entirely deceptive with regard to 

 the ancestry of vertebrates. 



Zoologists have not quite made up their minds 



