86 
ON THE BREATHING APPARATUS 
like Calopteryx, the Branchial plates are enclosed in an internal 
sac, hut the three large plates are replaced by about three hundred 
much smaller ones, which are arranged most symmetrically 
in six rows ; each row having the appearance of being double, 
owing to the plates laying first on one side and next on the 
other, giving to each row very much the appearance of an 
acacia leaf with its leaflets. 
These rows are connected with the general trachea by quite 
an extensive system of small branches. First, there are two 
tubes which supply the intestines, &c. ; these are connected with 
the two lower rows of hranchicB, each tube running the whole 
length of a row, and throwing off branches en route, but upon 
the outside of the sac. Then there are the two large trachea 
running the whole length of the body, and supplying head, legs, 
and embryo wings ; these are connected with the four upper 
rows of j)lates, each tube passing down outside two rows of 
plates and sending out branches to the individual hranchice. 
These branches, on entering the plates, suh-divide into the 
minutest threads which entirely fill the space between the thin 
walls, and thus offer a comparatively enormous surface for 
aeration. 
In jEseJma the same apparatus is present in all its details, 
the only difference being in the shape of the plates. This sac 
is closed by three semicircular plates, closing partly over each 
other, and situated in the extremity of the abdomen, being 
protected by five pointed plates which open and close at will, 
and assist in making the propulsive jet more effectual by 
concentrating it. 
These five pointed plates are situated on the ninth segment 
of the abdomen, and, on finding this was the case, I wondered 
considerably what had become of the other two segments, which 
are present in Agrion. 
Now, if w^e return to the above remarks on Agrion, it will be 
