CRANGON VULGARIS. 11 



first its axis is directed obliquely backward but this is soon 

 changed, apparently by a more rapid growth of mesoderm 

 to a nearly similar angle toward the anterior end of the* 

 embryo. 



From this point until stage G the changes are compara- 

 tively slight and are shown in figs. 37, 43 and 42 which rep- 

 resent longitudinal sections of stages E and G respectively. 

 The former does not show the inner extremities of either 

 stomodeum or proctodeum, the section being slightly ob- 

 lique but in the latter (fig. 42) both are well shown. In 

 both, the stomodeum has become bent, the ventral half being 

 directed upwards and forwards while the inner portion is 

 directed as strongly in the opposite direction. In my 

 sections it appears to abut abruptly against the mesoderm 

 though I cannot state positively that it does. 



The proctodeum on the other hand exhibits more marked 

 changes. ' The anus in fig. 43 is near the tip of the abdo- 

 men, but it is to be noted that it has lost its dorsal posi- 

 tion and has appeared upon the ventral surface, a feature 

 which is more strongly marked in fig. 42. Its inner ex- 

 tremity forms a wide funnel, the open mouth of which em- 

 braces the yolk, which now is seen to be breaking up (vide 

 fig. 20 and, later, fig. 23), the yolk granules and balls 

 circulating with a peristaltic motion in the intestine. My 

 series of sections seem clearly to show that the whole of 

 this intestine is of ectodermal origin and that the entoderm 

 has nothing to do with its formation. In stage If, fig. 54, 

 the same relations of the proctodeum to the yolk mass are 

 seen, and in the living embryo the movements of the yolk 

 particles are much stronger. 



In the same stage (//), the stomodeal division of the 

 alimentary tract exhibited some notable changes. At its 

 external extremity, it has become widened out to form a 

 buccal cavity (fig. 238, be) which abruptly contracts to 



