OF ARTS AND SCIENCES. 73 



Secondary Caudal Vesicles. — Soon after Kupffer's vesicle begins to 

 decline, we generally find a variable number of much smaller vesicles 

 making their appearance between it and the hind end of the embryo. 

 These secondary vesicles lie beneath the embryo in a somewhat 

 thickened portion of the periblast. These grow larger, probably by 

 coalescing, and may often be seen just in front of the terminal de- 

 scending portion of the alimentary canal, until the tail attains a length 

 of a millimeter or more. "Whether the contents of these vesicles may 

 be regarded as identical with the contents of Kupffer's vesicle, or 

 whether any genetic relation exists between the two, we are not pre- 

 pared to say. The general appearance of the two classes of vesicles 

 is the same ; and, until we had traced the disappearance of Kupffer's 

 vesicle, we naturally assumed that the secondary vesicles arose by 

 division of the primary vesicle, or that several primary vesicles co- 

 existed from the outset. 



Neurenteric Canal. — A surface view of the nearly closed blastopore 

 shows that the epidermal cells surrounding it are much elongated in a 

 radial direction. If the embryo be so placed that the remnant of the 

 blastopore is seen obliquely, it will be seen to form a funnel-like de- 

 pression, from the deeper and narrower portion of which a more or 

 less distinct streak may be traced completely through this part of 

 the embryo, which terminates at or near the posterior boundary of 

 Kupffer's vesicle. In some cases this streak presents the form of a 

 linear canal bounded by epithelium-like cells. The lumen of the 

 canal is reduced to a mere line. It is difficult to recognize distinct 

 boundaries to the wall of the canal in living embryos ; and .our sec- 

 tions have not thus far given us a satisfactory view of its relations to 

 tlie alimentary tract. As the caudal plate (Ryder) thickens up, the 

 canal, or streak representing it, appears to travel backward, which may 

 be explained by supposing that the portion of the ring lying behind it 

 is actually carried forward in order to form the hind end of the em- 

 bryo. That such a migration of cells takes place is not absolutely 

 certain ; but the evidence is in favor of it, and the previous relation 

 of the ring to the embryo sustains this view. There is at no time any 

 nearer approach to a true neurenteric canal than we have described. 



Tlie Formation of the Embryo. — Our conclusions respecting the 

 nature of the process by which the embryonic ring becomes converted 

 into the embryo are essentially the same as those of His and Rauber. 

 The so-called differentiation theory of Balfour and others fails to give 

 any satisfactory account of the relation of the two lateral halves of 

 the embryo to the ring, and offers no explanation of the "marginal 



