280 Mr. H. McE. Knower on the 



the " uncler-layer " follows the growth of the ectoderm, which 

 is somewhat more rapid. A sagittal section of a band before 

 segmentation shows the " under-layer " as a single row of 

 cells beneath the anterior portions of the ectoderm, where its 

 cells are sharply marked oif from the ectoderm. Beneath the 

 tail-end of such an embryo the " under-layer " cells are 

 collected into a large mass, which is not sharply separated 

 from the ectoderm. Most of the extension of the " lower- 

 layer " is apparently due to the multiplication of its own 

 cells. I have not yet studied the differentiation of this layer, 

 but can state that the endoderm appears after the establishment 

 of segments, and is not formed from vitellophags. 



The sections just referred to show well the changes of the 

 amnion until it has become a thin lamella. This is brought 

 about as a result of the anterior and posterior extension of the 

 embryo. The cells of the amnion are pulled out into a single 

 row anteriorly, while posteriorly for some time the membrane 

 retains something of its early appearance. When the tail- 

 end of the germ-band has reached the anterior pole of the 

 egg in its elongation, the abdominal region sinks gradually 

 into the yolk and the posterior extremity coils over ventrally 

 toward the head, giving the embryo an 8 shape. The ap- 

 pendages have meanwhile grown to nearly their definitive 

 length. The first and second maxillas are trilobed, and ten 

 rather prominent rudimentary abdominal appendages have 

 appeared. The cephalic region has changed considerably. 

 Just after the appearance of the appendages the lateral 

 margins of the cephalic lobes began to roll up toward the 

 mid-dorsal line. As this process continued a little pocket 

 was formed on either side of the head, which grew gradually 

 larger as the folds of the cephalic lobes approached the median 

 dorsal line. Finally the two lateral pouches fused to form 

 the head-cavity. The antennse were included in this folding, 

 and hence now enclose a portion of the head-cavity. 



When the embryo has reached the stage just described it 

 resembles Brandt's figure 11 of Calopteryx (1), but it is not 

 " immersed " in the yolk. It is impossible to say just how 

 long this " inverted " position is maintained. There is, 

 perhaps, a rather short interval before " revolution," which is 

 accomplished as described by Brandt for the Libellulid (that 

 is, judging from preserved specimens). When " revolution " 

 is over the embryo lies with its head at the anterior pole of 

 the eggj while the tail-end lies beneath the micropyles at the 

 posterior pole. The ventral surface of the embryo, as in early 

 stages, lies on the micropylar side of the e.gg. 



I have not studied the development beyond the appear- 



