Marshall—On the Anatomy of the Dragonfly. 773 
uninuclear -cells, the boundaries of which are difficult to dis¬ 
tinguish (fig. 34). 
Many of the acini look as if they were hollow, the central 
part containing a secretion. This appearance is due to the 
difference in structure in different parts of the cells. In the 
basal half or two-thirds of each cell (the nucleus is always with¬ 
in this part) the cytoplasm stains readily; in contrast to this 
the outer half or one-third of each cell did not stain very dark 
with haematoxylin. In the acini this darker-staining basal 
portion appeared as the wall and the lighter part as the cen¬ 
tral cavity. Upon closer examination one can see both these- 
parts belong to the cells and that the portion in which the 
stain was not so effective was dne to the accumulation of se¬ 
cretion in the outer half or third of the cell. 
Both Poletaew (19) and Bor das (3) speak of the swollen 
parts of the ducts situated on the base of the hypopharynx as 
the reservoirs. Bordas (2) in his first paper does not use this 
word but calls them “glandes buccales”. Poletaiew (19) in 
the quotation we have already given at the beginning of the ac¬ 
count of the salivary glands, used the word reservoir; he de¬ 
scribes the peculiar external appearance of the reservoirs as due 
to the euticular layer and says: “die an ein netzformiges Zel- 
lengeflecht aus dem Gewebe einiger Blatter erinnert.” In plate 
1, figure 5, he shows the reservoirs of a species of Lestes, the 
markings on which, as he gives it, resemble the shingles on a 
house, the shingles having rounded, exposed ends. Bordas (3) 
figures the reservoirs in three species of dragonflies ; Agrion 
minium ■, Cordulegaster annulatus and Cordulina aenea. All of 
these appear very similar to what has been found in Libellula 
4-maculata. Bordas gives enlarged views of a part of a res¬ 
ervoir of Anax formosa, both from the surface and in section. 
In his surface view the cells are shown with a polygonal outline, 
in section the wall is shown to be equally thick throughout 
and without the large protrusions so noticeable in the species 
here described. Berlese (1) in his text-book figures three cells, 
in section, from the reservoir of a species of Diplax and labels 
