346 KEPOET— 1892. 



described as a hypertrophy of the anterior end of the notochord, in that 

 it either extended unusaally far forwards, or, as was more frequently the 

 case, it bent round sharjjly on itself in front. 



The tail arises as a solid outgrowth from the body of the larva, con- 

 taining the greater part of the notochord, some mesodermic elements in 

 the form of two lateral plates of cells, and some endoderm in the shape of 

 a median ventral cord of cells, two deep. 



The intestine of Ascidians is in all probability entirely homologous 

 with that of Ampldoxus, including the possession of a caecum ; and its 

 U-shaped curvature is appai-ently, in accordance with the reflections of 

 Lang, to be correlated with the sessile habit of the Ascidians. An in- 

 teresting parallel for this is afforded by the comparison of Cephalodiscus 

 and Balanoglossus. 



Being unable to rear the embryos of Gynthia papillosa beyond the tad- 

 pole stage, I commenced to work at the post-embryonic development of 

 Ciona intestinalis. I have already sent up a short resume of part of my 

 researches on this form, together with some which I instituted on 

 Clavelina lepadiformis for purposes of comparison, for publication in the 

 ' Proceedings of the Royal Society.' The paper in question was read 

 before the Royal Society on May 18 of this year (1892). It contained 

 an account of what I called a prohoscis cavity, which, on account of its 

 primary relative position to the endostyle, I homologised with the head 

 cavity or prse-oral coelom, or, in short, proboscis cavity of Amphioxus taken 

 in conjunction with the prse-oral pit. 



In Giona as in Amphioxus the primary axis of the endostyle is at right 

 angles to its deflnitive axis, and it lies in both cases tout-d-fait anterior — 

 that is to say, immediately behind the cavity of the proboscis. The latter 

 structure in Giona carries at its anterior extremity the three adhering 

 papillfe. 



The first four primai-y stigmata of Giona arise on each side by sub- 

 division of one gill-slit, and not, as van Beneden and Julin assumed in 

 Phallusia scahro'ides, by separate perforations. This gill-slit I have 

 called the true first gill-slit of Ascidians, and homologised the pair of them 

 with the first pair of slits of Amphioxus which I have previously described 

 as closing up in the course of the metamorphosis. 



In Giona the heart arises by the splitting apart of the two layers com- 

 posing the septum, which at first divides the pericardium into two halves, 

 while in Glavelina, as shown by van Beneden and Julin, the correspond- 

 ing septum breaks down, and the heart forms by invagination of the 

 dorsal wall of the pericardium. 



In both Glavelina and Giona the hypophysis is at first in free com- 

 munication with the cerebral vesicle ; it is, in fact, largely formed by con- 

 striction from the wall of this vesicle. This is in profound opposition to 

 the account of van Beneden and Jalin, who state that in Glavelina the 

 hypophysis arises entirely by a caecal outgrowth from the wall of the 

 branchial cavity. I am absolutely at a loss to explain how they came to 

 this result. My observations were first of all made on the larvas of 

 Giona, where the organs are extremely small and difficult to interpret ; but 

 still I convinced myself of the true condition of things, and immediately 

 assumed that in Glavelina the formation of the hypophysis occurred in & 

 rather different manner, which could be explained by the great modifica- 

 tion which Glavelina shows in its general development. When, however, 

 I came to examine how matters stood in the larva of Glavelina, I saw at 



