SECT. 4] HEAT-PRODUCTION OF THE EMBRYO 673 



could go in the absence of a circulation. Bataillon pointed out that 

 this temporary cessation of mitotic activity corresponded exactly with 

 the trough in carbon dioxide production shown for the teleost in 

 Fig. 124, between the 40th and 50th hours. The mitotic divisions 

 having localised themselves round the edge of the embryonic area, 

 the central part becomes dislocated, or rather raised up, and the 

 deep cells, many of which show amoeboid processes, migrate to the 

 periphery. Between the raised up germinal area and the periblastic 

 layer a fluid appears, thus making a very irregular segmentation 

 cavity, in which the deep-lying cells, finding better conditions, begin 

 to proHferate and form the primitive endoderm. These cellular 

 displacements and migrations, these mechanical difficulties, due to 

 the beginning of extension of surface, said Bataillon, are probably 

 the obstacles which give to that particular stage in teleostean develop- 

 ment its peculiar characteristics, among the most striking of which is 

 the trough in the carbon dioxide production. Then, after this stage, 

 the extension of the blastoderm begun by the proliferation of the 

 cells at the edges of the germinal area goes on continuously until 

 the yolk is completely covered. In the case of the amphibian embryo, 

 none of these difficulties arise, for the extension of the ectodermal 

 elements at the animal pole over the yolk-laden cells of the vegetal 

 pole is a relatively simple process. 



In commenting on Bataillon's investigations, the resemblance be- 

 tween his curve and that of Hyman must be carefully considered, 

 for, although the time correspondence is sufficiently close to warrant 

 the belief in a real agreement, the descriptions given by the two 

 writers are slightly at variance. Phonixus, according to Bataillon, 

 accomplishes its surrounding of the yolk by the blastoderm after 

 the great trough, but Fundulus, according to Hyman, does it partly 

 before. Again, the curve of Hyman, though of the same shape as 

 Bataillon's, is more lengthened out along the time axis, so that the 

 peaks and troughs do not exactly correspond. Balfour, moreover, does 

 not describe the static condition on which Bataillon lays such stress. It 

 is therefore difficult to appraise Bataillon's results. As regards the theo- 

 retical side of his work, it might perhaps be observed that his simple 

 direct correlation between amount of carbon dioxide put out per hour 

 and number of mitoses going on is not altogether satisfactory. We have 

 no evidence that there is a general connection between these two 

 phenomena. Then one might ask why cell-migration should not be 



