214 Mr. J. T. Cunningliam on some 



surface of the oil-globule. These pigment-cells must have 

 been derived from the periblast, and afford an undeniable 

 example of mesoblastic cells formed directly from that layer. 

 The occurrence and significance of these chromatophores was, 

 I believe, first pointed out by John A. Ryder (17), in the 

 'American Naturalist,' Extra, November 1886; they have 

 also been described and discussed by myself in my paper in 

 the Journ. Mar. Biol. Assoc. (4). In the same paper I 

 described the occurrence of black chromatophores at the surface 

 of the periblast in the embryo of Pleuronectes microcephalus ; 

 these pigment-cells are also developed directly from tlie peri- 

 blast, no extension of the lateral mesoblast over the yolk 

 having taken place. 



Finally, it has been maintained by Ryder and others that 

 colourless blood-cells are segmented off" from the periblast and 

 enter the venous sinus and heart. I have not observed this 

 myself; but there seems to be good evidence for its occur- 

 rence. 



Thus the periblast with the yolk is evidently equivalent to 

 the yolk-cells of the Amphibian ovum, and, like those cells, 

 continues throughout development to take part in the segmen- 

 tation, though in a different fashion, a part of the periblast 

 remaining unsegmented, although nucleated, until the final 

 absorption of the yolk. Both the yolk-cells and the periblast 

 may be termed after a certain stage primitive hypoblast, 

 provided it be remembered that a great part of the primitive 

 hypoblast gives rise to mesoblastic cells. This last fact is 

 well brought out by Shipley in his paper " On the Develop- 

 ment of I'etromyzoriy^ where he says : — " The first formation 

 of the mesoblastic plates appears to take place by a diff'eren- 

 tiation of the hypoblastic yolk- cells in situ, and not from 

 invaginated cells." This differentiation consists chiefly in 

 subdivision, by Avhich small cells of the mesoblast are pro- 

 duced from the large yolk-cells. In fact a little consideration 

 shows that the segmentation of cells from the Teleostean 

 periblast to form hypoblast and mesoblastic tissues corre- 

 sponds perfectly Avith the subdivision of the yolk-cells in 

 I'etromyzon and Amphibians which gives rise to hypoblast 

 and mesoblast in those forms. 



Relation of Oil-glohules to Periblast. 



As 1 have already said, in the mature undeveloped ovum oil- 

 globules are sometimes free in the yolk, sometimes fixed by en- 

 velopment in the cortical protoplasm of the egg. But in all cases, 

 so lar as my experience goes, the oil-globules are during deve- 



