THOMPSON: ANATOMY OF MOSQUITO. 
165 
the sacs, and that the pouches have an aerostatic function. Nuttall 
and Shipley (: 01-: 03) have thoroughly exposed the inaccuracy of this 
account of the sacs of Anopheles. It would be equally inapplicable 
to Culex. The two genera show no essential differences with 
respect to the structure of the food reservoirs. 
Salivary glands .— The epithelium of the salivary duct is thin 
but columnar and the duct has a trachea-like aspect owing to annu¬ 
lar thickenings of its intima. It forks within the head and in the 
thorax each branch supplies a tripartite gland. The duct within 
each of the three aciui of the gland is (Cidex stimulans) of uniform 
diameter, has a delicate smooth intima, and is traceable to the apex 
of the acinus where it ends blindly. The acini are elongate and as 
their first describer Macloskie (’88) and later writers have noted, 
the central and lateral divisions are unlike in appearance. Maclos¬ 
kie gives no evidence to support his suggestion that the central 
acinus secretes the poisonous element of the saliva. This acinus 
(pi. 14, fig. 23, ca) is slender, cylindrical, and its nuclei are promi¬ 
nent, surrounded by granular cell contents. In sections the larger 
part of each cell is filled with secretion which, as precipitated by 
the reagents, stains lightly with haematoxylin and is coarsely granu¬ 
lar. The nuclei are pushed over against the basement membrane. 
The lateral acini apically are of the same diameter as the central 
acinus, but their bases are distended. Their cells are filled with a 
transparent secretion of high refractive index. Toward the base of 
the gland nearly every cell contains a huge ovoid vacuole filled with 
the secretion which forces the nucleus against the basement mem¬ 
brane. Sections show clear spaces where these vacuoles were situ¬ 
ated. Nearer the apices of the acini the vacuoles are smaller and 
less numerous. 
The differences in the appearance of the lateral acini and the cen¬ 
tral acinus are due to the different substances secreted and stored 
in the cells. During the later pupa stage the distinctions are want¬ 
ing and all parts of the gland resemble the central acinus of the 
imago as viewed by transmitted light. Sections, however, show 
no secretion in the cells. The differences make their appearance a 
short time before the imago emerges. 
The midgut .— The part of the midgut which receives the eso¬ 
phageal valve (pi. 14, fig. 20) is best called the cardia. As with 
most other flies, the region is not definitely marked off from the 
