8 
Proceedings of the Royal Society 
the crystalline lens, none of these arises in the Teleostean by direct 
invagination from the exterior, as they all do in Elasmobranchs. 
The neurenteric canal, as a distinct lumen, does not exist in Teleos- 
tean embryos, nor does the blastopore which would give rise to it. 
In Petromyzon we have an interesting condition intermediate 
between the Elasmobranch and the Teleostean. The neural canal 
does not arise as a groove open to the exterior, but, as in Teleosteans 
appears at a later stage closed from the beginning; while the 
gastrula cavity, on the other hand, is formed by direct invagination. 
The part of the blastopore represented by the neurenteric canal has 
disappeared, and not the part which in the Elasmobranch remains 
open to the exterior longest (see Plate I. figs. 2, 3 ; Plate II. fig. 5). 
As far as actual comparison goes, therefore, there is every reason 
to believe that Kupffer’s vesicle represents the gastrula cavity, or 
that part of it which is left after deducting the body cavity ; but still 
further evidence is offered by the subsequent history of the vesicle. 
In Plate II. fig. 4 is shown the condition of the herring embryo on 
the fourth day in the region where Kupffer’s vesicle existed. The 
intestine is completely formed ; and it will be noticed that beneath 
the intestine the periblast is yery thin, and contains no nuclei, 
though these are visible laterally beneath the mesoblast. It seems 
clear, then, that the floor of the intestine in this region has been 
formed by the differentiation of cells round the nuclei of the peri- 
blast, exactly as in Elasmobranchs the periblastic floor of the gas- 
trula cavity is transformed into the floor of the intestine. 
I think there is no room for doubt that the significance of 
Kupffer’s vesicle is completely elucidated by the facts and com- 
parisons I have thus given ; it is the last rudiment of the invagina- 
tion cavity in the Teleostean. 
Remarks on General Vertebrate Morphology . — The view which 
obtains homologies between Vertebrates on the one hand, and Chae- 
topoda, Crustacea, and Insecta on the other, by inverting the latter, 
was suggested long ago by Geoffrey St Hilaire, when he said that 
Vertebrates were insects walking on their backs. But the distinct 
hypothesis that a Vertebrate is actually descended from an ancestor 
which agreed in almost every point of its anatomy with a modem 
Chaetopod, owes its origin and support in great measure to Dr Dohrn, 
director of the Zoological Station at Naples. St Hilaire’s suggestion 
