764 Wisconsin Academy of Sciences, Arts, and Letters. 
eight, the medium dorsal fold is still larger than any of the 
others. Between these large folds there are still many small 
irregular ones. At the middle of the cervical region a change 
takes place; the large median dorsal fold persists and at either 
side of this there are two or three narrow ones ventrally di¬ 
rected. On the ventral wall are a number of small folds (fig. 
13). Further hack in the cervical region the folding again 
changes and one could not, in this region, determine in a trans¬ 
verse section, which part was dorsal: the median dorsal fold 
is still present hut no larger than the others. Throughout 
this and other parts of the fore-intestine the cuticular layer 
is covered wih small blunt processes (fig. 14) ; similar ones 
were noted as present in the pharynx. 
In the middle of the prothorax the oesophagus enlarges; 
the walls remain of the same thickness but, owing to the in¬ 
crease in diameter, the folds do not come so close together. 
Towards the posterior end of the prothorax the large folds dis¬ 
appear and their place is taken by a number of small ones. 
As the oesophagus enters the mesothorax there is, 4.25mm 
from its commencement, the sudden increase in its diameter, 
due to the abrupt fall in its ventral wall. At the beginning 
of this enlarged part the folds are smaller and most of them 
are restricted to the dorsal wall. At this region the wall of 
tthe oesophagus becomes thinner, due to a decrease in the layer 
of circular muscles (fig. 15) which are not so thick here as in 
the most anterior region. From this place to the gizzard the 
muscular layers are scarcely noticeable. 
Gizzard. If a specimen of Libellula with an empty crop be 
examined the plates of the gizzard can be seen through the thin 
wall protruding, apparently, from the anterior end of the 
mid-intestine forward into the crop. One is unable to notice 
any boundary between the crop and the gizzard. If the in¬ 
testine at this region be cut longitudinally into halves a view 
of the interior will show the true position of the plates and 
it will then be seen that they have no connection with the mid¬ 
intestine (fig. 16). Taking for the limits of the gizzard a 
point slightly anterior to the plates and, posteriorly, where the 
