DEVELOPMENT OF TEACHIKTJS VIPEKA. 281 



rim as it extends over tlie yolk, until the latter is completely 

 enclosed. The application of reagents, by bringing about a 

 shrinking of the yolk, causes the outer membrane to separate 

 further from the inner or intermediary layer which remains 

 attached to the yolk. This shows that a fissure still exists 

 between the two layers ; indeed it can be traced all through the 

 embryonic life so long as the yolk-sac persists. The distance 

 between the two membranes of this cavity becomes more pro- 

 nounced in the newly hatched embryo (fig. 27). I have also, 

 as shown in fig. 14, traced a prolongation of the segmentation- 

 cavity beneath the fore part of the head ; but have not satisfied 

 myself that it is pushed forward to form the pericardiac cavity, 

 as is maintained by Ryder ; although later stages seem to support 

 this view. If the cavity really does extend so far, the line formed 

 by its two membranes under the hind portion of the head is : o 

 minute in the living egg as to look single only. In the newly 

 hatched embryo, however, part of the yolk has been absorbed, 

 and the heart then shows clearly in the same cavity as is continued 

 around the yolk. 



Kingsley and Conn, in the text of their paper, combat Eyder's 

 views on this point altogether, but add, in a footnote, that later 

 studies make them more inclined to accept them, at any rate 

 partially. 



The first step towards the formation of the embryo is the 

 appearance of a faint streak, the keel, in the median line of 

 the anterior portion of the embryonal shield. This keel 

 lengthens, and becomes club-shaped as the embryonal area 

 encroaches on the segmentation-cavity. As the development 

 advances, the anterior portion of the keel is seen to widen, and 

 the outline becomes spathulate. This is caused by outgrowths 

 springing from the keel, one on each side, to form the optic lobes. 

 At first these only appear as slight swellings, but soon a curved 

 line makes its appearance on each side, marking the angle formed 

 by the optic lobes with the keel. By this time the growing 

 blastoderm has already passed the equator of the egg. The optic 

 lobes begin to be segmented off before the protovertebrse form, 

 and the process is completed about the time four or five proto- 

 vertebrse are visible. The lumen begins to show before eight 

 protovertebrse are formed, and before the blastodermic ring has 

 closed. 



The division of the fore part of the medullary cord into fore 



