232 WILLIAM BATESON. 



this larva, mainly in its extent, and in the fact that in Amphioxus 

 it encloses a neurenteric canal. Now, with regard to the second 

 of these differences, from a consideration of the position of the 

 blastopore in my larva (fig. 25), it is clearly only necessary to 

 imagine the invagination of the dorsal nerve cord to have been 

 extended along the back (instead of being confined to the region 

 of the collar) in order to reproduce the condition which is 

 found in Amphioxus ; for the dorsal nerve cord in the adult 

 Balanoglossus is as a matter of fact continued into this region, 

 though in connection with the skin. 



With regard to the origin of the mesoblast, the following is 

 the arrangement in Amphioxus (Hatschek). It is formed 

 anteriorly by a primitively unpaired pouch of hypoblast, which 

 is continued into two posterior horns ; this anterior pouch is 

 followed by a great number of paired pouches lying on each 

 side of the body which are constricted off from the gut. The 

 anterior pouch is the last to close. As it does so, its cavity 

 divides into a pair of pouches, lying one on the right hand, the 

 other on the left. Of these the cells of the left become 

 columnar and ciliated, and its cavity opens to the exterior, 

 while the tissue of the right pouch becomes flattened epi- 

 thelium, lining the body cavity of the head. On the other 

 hand, in the larva now described the mesoblast is formed from 

 an anterior archenteric pouch with two posterior horns, followed 

 by only two pairs of pouches. Of the two incompletely sepa- 

 rated divisions of the anterior cavity, that which lies on the 

 left side becomes lined by ciliated columnar epithelium, and 

 opens to the exterior, while the right hand one forms con- 

 nective tissue. The origin of the mesoblast in Amphioxus 

 differs, therefore, in being accomplished by a great number 

 of paired posterior pouches instead of by two ; and in the 

 fact that the division between the right and left parts of the 

 anterior pouch is completed instead of being partial. There 

 appears, therefore, to be a general agreement in the early 

 development of these two animals which holds good even in 

 the remarkable asymmetry above described. 



At first sight it seems likely that these points of resem- 



