636 CIRCULATION OF THE BLOOD AND DORSAL VESSEL. 
understood the function of the dorsal vessel, Cuvier denied the 
existence of a circulation in Insects. C.G. Carus [807] made 
many observations on this phenomenon, and clearly established 
the fact that not only the blood circulates, but in many Insects, 
perhaps in all, the course of the hzmal fluid is circumscribed 
by very definite channels. 
Indeed, long before, Baker [806] in 1755 described not only 
the pulsations of the dorsal vessel, but observed the blood 
flowing in very definite channels in the wing of a species of 
Locust ; and in 1837 Bowerbank [810] gave a detailed account 
of the circulation in the wings of Chrysopa perla. He describes 
the blood current as flowing steadily through all the nervures 
towards the apex of the wing, and as returning by a single 
channel, running in the posterior margin of the wing, in a 
more rapid stream, to the thorax. 
Definite Course of the Blood in the Appendages.—Verloren tried 
to account for the circulation in the wings and legs by the 
supposed existence of small accessory hearts, similar to lymph 
hearts, in each appendage, but he had no evidence to offer in 
favour of the view, and to this day no such hearts have been 
discovered. In aquatic larve the circulation in the leaf-like 
gills and in the anal sete is most rapid, and it is altogether 
similar to that in the legs, wings and elytra of perfect insects. 
In the former, lymph hearts could not fail to be seen if they 
existed. 
The great cavities of the body of the imaginal insect are 
divided from each other by narrowings, the small openings in 
the diaphragmata, and the narrow abdominal pedicle of the 
Wasp, for example ; and not only transverse, but longitudinal 
more or less complete septa traverse the body. Similar 
septa also exist in the appendages. Thus the blood flows 
into the head above the tentorium, and leaves it below 
the tentorium; and the circulation through the proboscis is 
provided for by blood sinuses on its dorsal surface which 
communicate with the region above the tentorium, and 
on its ventral surface with the region below it; these 
intercommunicate by narrow channels offering considerable 
