350 TUNICATA. 



region corresponding to the future tail of the larva. In the anterior, 

 dilated region, the mesoderm arises through the development of 

 paired diverticula of the archenteron (Figs. 159, A, ms), the lumens 

 of these sacs very soon disappear and the cells of their walls, 

 which originally were arranged in two layers (the somatic and the 

 splanchnic layers), then appear irregularly distributed between the 

 ectoderm and the entoderm. Between the two coelomic diverticula, 

 the roof the archenteric wall is completed by entoderm-cells (Figs. 

 159, 160 A, 161 A, ch) which represent the rudiment of the chorda. 

 At a later stage, these cells become shifted into closer proximity, and 

 thus form a solid strand which, in cross-section, is round. The fact 

 that this strand is to be traced back to a median fold of the arch- 

 enteron is proved by transverse sections through the most anterior 

 part of the rudiment of the chorda (Figs. 160 A, 161 A), where the 

 infolding of the cell -plate which represents the rudiment of the 

 chorda can actually be seen (VAN BENEDEN and JULIN, No. 10). 



In Amphioxus, according to HATSCHEK, the median fold of the intestinal 

 wall is not completely absorbed in the formation of the chorda. Its lateral 

 cells, which are in contact with the coelomic diverticula, yield the cells which 

 complete the dorsal wall of the intestine after the chorda has separated from 

 the mesoderm. VAN BENEDEN and JULIN conjecture that similar conditions 

 exist in the anterior part of the chorda-rudiment in the Ascidians (cf. Fig. 

 161 A, x). 



A certain guarantee of the accuracy of the observations made by VAN 

 BENEDEN and JULIN seems to be afforded by the striking resemblance to the 

 formation of the mesoderm in Amphioxus. Their observations, however, 

 have not been confirmed either by DAVIDOFF (No. 14), who investigated the 

 formation of the mesoderm in Clavelina and Distaplia or by WILLEY (No. 

 54a). According to the former author, the mesoderm-cells become abstricted 

 from definite entoderm-cells at the margin of the blastopore (mesoderm- 

 gonads), and become arranged in an originally single layer below the 

 entoderm. This would be a kind of mesoderm-formation by delamination. 

 DAVIDOFF was unable to find coelomic diverticula. The " mesoderm-gonads " 

 are said to remain in connection with the entoderm, and, after the mesoderm 

 has been produced, some of them are said to take part in the formation of the 

 chorda and others in that of the alimentary canal. * 



In the posterior region of the body (the future caudal region), the 

 condition appears to be modified through the early disappearance of 

 the lumen of the intestine (Fig. 160 (7). The sl5rand-like chorda is 



* [In Ciona, the mesoderm-cells form temporarily part of the wall of the 

 archenteron between the chorda and the entoderm. Eventually they become 

 displaced outwardly and the entoderm and chorda come into contact. Ac- 

 cording to CASTLE (No. II.) there does not appear to be any enterocoelic 

 formation in this genus. ED.] 



