216 Mr. J. T. Cunnindiara on some 



O' 



in fact in all eggs wliicli exhibit vitelline vessels, the segraen- 

 tation-cavitj is obliterated after the closure of the blastopore 

 by the production of mesoblast between the epiblast and 

 periblast. This mesoblast is produced partly by the extension 

 of the mesoblastic layer from the lateral region of the dorsal 

 embryonic rudiment, partly, I believe, by the formation of 

 mesoblastic cells from the periblast. In the mesoblast thus 

 produced around the yolk, tubes are hollowed out to form the 

 vitelline arteries and veins, the veins becoming continuous 

 with the cavity of the heart, which is formed in a similar way 

 in the mesoblast ventral to the pharynx. 



In pelagic ova and in the ova of the herring the course of 

 affairs is somewhat different. In these ova and the larvae 

 hatched from them there are no vitelline blood-vessels. In 

 them the segmentation-cavity may, and probably does, dis- 

 appear for a time after the closure of the blastopore in conse- 

 quence of its epiblastic and periblastic walls coming into 

 contact. But the lateral mesoblastic plates do not extend into 

 it and obliterate it. When the heart commences to be formed 

 the segmentation-cavity seems to reappear; that is to say, a 

 cavity appears between the periblast and the epiblast of the 

 yolk-sac. This cavity is continuous all round the ventral 

 region and sides of the yolk, and anteriorly it is in communi- 

 cation by a definite large aperture with the posterior end of 

 the auricle of the heart. But this cavity is theoretically no 

 longer the segmentation-cavity ; it is, at least on its inner or 

 periblastic side, partially lined by mesoblastic cells, namely 

 chromatophores produced from the periblast. Morphologically, 

 as I have pointed out in my paper in the ' Journal of the 

 Marine Biological Association ' (4) , this cavity is homolo- 

 gous with the vitelline blood-vessels in the salmon embryo, 

 and, like those vessels, it is continuous with the auricle of 

 the heart. It is shut off from the pericardium by a definite 

 continuous mesoblastic membrane, and it is also completely 

 separated from the body- cavity formed in the mesoblast at 

 the sides of the embryo. It is not till a late stage of develop- 

 ment, namely when the yolk has been entirely absorbed, that 

 the mesoblast is sufficiently developed ventrally to divide 

 up this perivitelline blood-sinus into separate blood-vessels, 

 the blood-vessels which in the adult form the veins and arteries 

 of the viscera. 



I claim the credit of having been the first to give tliis 

 explanation of the fact that in pelagic ova the heart is in open 

 communication posteriorly with a continuous cavity round 

 tlie yolk, a cavity wliich appears to be the segmentation- 

 cavity. Shipley (19) has shown that an exactly similar 



