158 ALICE M. BORING. 



the size of the piece or the part of the tube, and if the endoderm 

 and ectoderm both take part at one end, it is reasonable to expect 

 that they would along the entire edge. With this idea in mind, 

 the apparently solid closures were studied in order to see if there 

 were not two layers of endoderm. The pieces were traced back 

 through the series of sections to the place where less had been 

 cut off in order to locate the ridges and in many pieces, a ridge 

 was found included in the small piece. In some very small 

 pieces, not including any ridge, it was found that the edges, both 

 ectoderm and endoderm, had met to enclose a small but distinct 

 cavity. Fig. H shows the section just before the edges meet. In 

 other small pieces with no ridges the higher powers of the micro- 

 scope showed two distinct layers of endoderm, indicating that the 

 closure had been the same, but the piece was so flat and the 

 angle turned by the closing membrane was so sharp that the 

 cavity between the old and new walls was flattened out into a 

 mere crack and the endoderm of the closing membrane practically 

 touched the endoderm of the old wall (Fig. E). The presence of 

 a ridge would cause the obliteration of the cavity in a piece other- 

 wise large enough to leave a distinct cavity in closing. Fig. 

 D shows this on the right side. The ridge is further toward the 

 right, and consequently that side has no cavity. This section 

 shows a circular arrangement of the endoderm cells at the point 

 of turning, indicating a compression at their inner ends that 

 would justify the statement that the layer of endoderm has been 

 drawn around with the ectoderm. 



In apparent contradiction to the above statement that only 

 small pieces, or pieces including a ridge, close in solid, obscuring 

 the cavity, a few sections complying with neither of these condi- 

 tions show the ectoderm beginning to grow over the endoderm 

 and leaving no cavity. But here, as in the small pieces there are 

 two layers of endoderm in all cases where the ectoderm has over- 

 grown, showing that the endoderm is following the ectoderm 

 closely, and that a two-layered plate is really being formed even 

 if obscured by the pressure of the two walls together (Fig. I). 

 What the further development of such pieces would be, has not 

 been definitely determined, but it is not likely that they close in 

 to make a solid piece, for among the completely closed solid 



