MUSEUM OF COMPARATIVE ZOOLOGY. 211 



pronephric thickening. The dorsal portion of the expanded body cavity 

 is the pronephric cliamber. 



The question whether the somatopleural thickening described in 

 Stage I. be an early condition of the pronephric thickening is only to be 

 answered by considering the fate of the former. Behind somite IV. this 

 early thickening wholly disappears, and the one which is seen at a later 

 stage is an independent formation. This conclusion is justified by a com- 

 parison of Figure 4 (Plate I.), showing the somatic layer to be only 

 one cell thick in the posterior region of an embryo of the pi-esent stage, 

 with Figure 3, which shows a two to three layered somatopleure {so'plu.) in 

 a somewhat more anterior region of an embryo of the next younger stage. 

 In the region of somites II., III., and IV., however, the somatopleure 

 never wholly thins out ; but the thickening is here moulded into a more 

 definite form, and becomes the fundament of the pronephros. To my 

 mind, it is as if the mesoderm, in the process of becoming thinner, was 

 overtaken by the necessity of afibrding material for the formation of the 

 pronephros and duct, and, as a matter of physiological economy, used 

 for that purpose an accumulation of cells already present. Indeed, 

 from the form of the thickening in anterior portions of the embryo, I am 

 disposed to regard the differentiation of the pronephric thickening in this 

 sense as having begun already in Stage I. 



The corresponding series of frontal sections shows five well developed 

 protovertebrse, representing somites I.-V. (Plate II. Figs. 13, 14). A 

 mass of mesenchymatic tissue in front of somite I. is doubtless the rem- 

 nant of the rudimentary anterior pi'otovertebra observed in the series of 

 cross sections, and behind somite V. the differentiation of a sixth is 

 faintly indicated. Above the level of the lower border of the chorda the 

 protovertebrae are sharply marked off from one another, and the somatic 

 layer is relatively thin. Near their ventral margins, however, the suc- 

 cessive protovertebrse are in close contact, and the somatic layer shows a 

 pronounced lateral thickening (Fig. 13, eras, pr^ni^h.). 



On passing ventrally to the region of the lateral plates, the inter- 

 protovertebral constrictions vanish. Since frontal sections, however, do 

 not here cut the layer of mesoderm perpendicularly, certain sections in 

 the series show a distinctly segmented splanchnic layer, while the so- 

 matic thickening in the same frontal plane is unsegmented. Farther 

 ventral there are no traces of segmentation in either layer. Here the 

 splanchnopleiire {spVplu.) uniformly consists of a single layer throughout 

 its entire extent. The somatopleure facing the ganglion nodosum, and 

 also that in the posterior region, is thin ; but in the anterior portion of 



