278 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



It seems to me to be quite superfluous to form a new sub- 

 genus for this species, H. cingalense, although it does not 

 seem to fall exactly into any of Willey 's sub-genera. 

 In choosing generic characters it is well to look for those 

 which are at once very constant and easily made out. 

 Among the most constant of organs in any animal are 

 the gonads, and in the Cephalochorda they are most con- 

 spicuous when present. Their arrangement, therefore, 

 is a good character on which to classify the group. The 

 nietapleural folds are also very constant, and these 

 together form the two fundamental, most important and 

 most constant characters. They show two series, (1) 

 biserial gonads and symmetrical metapleura, i.e., the 

 metapleura of both sides ending just behind the anus, 

 and (2) uniserial gonads and asymmetrical metapleura, 

 i.e., the left metapleuron dying away just behind the 

 anus, while the right is continuous behind with the 

 ventral median fin. This gives us two genera. Other 

 characters such as the shape of the fins, and the number 

 and arrangement of the myotomes vary so much as to 

 be only of use in determining species, and then only 

 when taken in conjunction with other characters. I, 

 therefore, agree with Willey in dividing the group into 

 two genera, only. The first series above gives us the 

 genus Branchiostoma, and the second series the genus 

 Asymmetron. 



As I propose that sub-genera should be abolished 

 altogether, the name Asymmetron should be applied to 

 the second series in preference to Willey's Heteropleuron, 

 because of priority. Thus we have the group primarily 

 divided into the two genera Branchiostoma and 

 Asymmetron. 



Willey recognised five sub-genera. Three of these, 

 Dolichorliynchus, Paranvphioxus and Epiyonichthys have 



