258 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
which the first two pairs are near the anterior pole of the egg. The 
endoderm has now made its appearance and is in process of surround¬ 
ing the yolk. As has been demonstrated by Yoeltzkow (’ 89 ) and 
Graber (’ 89 ) in Musca, by Ideymons (’ 95 ) in Forficula and various 
orthopterous insects, and by Lecaillon (’ 98 ) in Coleoptera, so in 
Melophagus the endoderm is a derivative of the inner ends of the 
stomodeal and proctodeal invaginations. The first indication of it 
is a proliferation of cells at the inner end of the proctodeum. This 
extends itself forward, in the form of a single layer of epithelium, 
along the dorsal surface of the yolk (the proctodeum having a 
dorsal position) to the head-fold (Fig. 12), and also around the 
sides of the yolk toward the ventral side of the egg (Fig. 15, 
en'drm .). As is to be seen in Figure 10 ( prd .), the proctodeal 
invagination, when it first appears, is bounded on its inner end by 
the mesodermic layer of the tail-fold. The boundary between 
mesoderm and ectoderm in this region is always perfectly easy to 
determine, because of the very different character of the cells of the 
two germ-layers. As the proctodeal invagination increases in 
depth and the fundaments of the Malpighian tubules begin to 
appear, the mesodermic layer gradually becomes thinner, until it 
entirely disappears. The cells of the anterior (deep) portion of the 
ectodermic invagination then proliferate rapidly and give rise to 
endoderm (Fig. 12, eri'drm.). 
The proctodeum, since it appears earlier than the stomodeum, 
apparently gives rise to the greater part of the endoderm, — an 
observation which was also made by Graber (’ 89 ) on Musca. In 
the stage represented by Figure 12 the endodermal epithelium 
has extended forward from the proctoderm to the head-fold, and 
lies beneath the stomodeum, but without having fused with it. I 
do not believe, however, that the stomodeum fails to cooperate in 
the production of the endoderm. As shown in the figure last men¬ 
tioned, the stomodeum has just pushed its way through the meso¬ 
dermic layer, which at first bounds its inner surface ; in the stages 
succeeding this, as will be shown very soon, this inner surface fuses 
with the endoderm and apparently aids in its formation. 
No trace of coelomic sacs appears in the mesoderm at any 
time. 
