44 • K. MITSUKURI. 



mesoblast is «till C(3ntinuoiis with the gut-by poblust. This inward 

 extension of the gut-hypoblast seeni8 to l)e the signal for the separa- 

 tion of the three structures above meritioned, for very soon we come 

 to sections like Fig. 18, in which the chovàn-aiilage, the mesoblast, 

 and the gut-hypoblast have separated from one another. From this, 

 it is evident that the epithelium lying between the chorda- 

 anlage and the gut-hypoblast (the lightly shaded part in 

 Fig. 16) and representing the walls of thegut-diverticu- 

 lum, becomes a part of the mesoblast. — the same conclusion 

 as was reached in the case of Fig. 23. 



If we now trace the series forwards from Fig. 16 we find the 

 proliferated mesoblast mass (darkly shaded in the figures) growing 

 smaller and smaller (Figs. 15 and 14). A few sections forward, we 

 come to the appearances depicted in Fig. 13. In this, we find laterally 

 the gut-hypoblast consisting more or less clearly of a single layer of 

 columnar cells. As we advance towards the median line, this epithe- 

 lium suddenly stops at some distance from the chovdu^-anlage, and 

 between it and the latter structure ^ve find a stretch of epithelium 

 with more than one layer of cells. This part I have shaded lightly in 

 the figure. The increase of cell-strata in this part is possibly due to 

 the fact that cells here are dividing rapidly to give rise to the meso- 

 blast mass. The darkly shaded part (the proliferated mesoblast cells) 

 in this figure ought therefore perliaps to have been extended further 

 inwards. 



P\n'ther forwards, the chorda-((///r/^t', the mesoblast, and the gut- 

 liypoblast are all as yet unditferentiatcd, and the loAver layer consists 

 of a single layer of columnar cells throughout. 



It is evident from the foregoing description, that the lightly 

 shaded part of Fig. 13 is equivalent to the similarly shaded part of 

 Figs. 16 and 12, which in its turn is comparable to the walls of the 



