AQUATIC INSECTS IN NEW YORK STATE 307 
stems examined were found numerous examples of larvae with 
their caudal spines pushed into the tissue of the plant clear 
up to their bases. In every case noted the larvae were at the 
apex of vigorously growing stems. As was described above, 
the larva eat large round holes in the apices of the growing 
stems, and this gnawing of the holes is undoubtedly done while 
the larva is respiring, as I have observed many larvae in the 
characteristic attitudes shown on plate 28, figure 1, and plate 
27, figure 18. These figures, I think, show the larvae in a 
frightened attitude, which resulted from pulling the plant from 
the mud, and, as a result, one larva stopped feeding while 
the other stopped respiring. 
Every investigator who has studied the function and struc- 
ture of the caudal spines has arrived at a different conclusion. 
Perris, who studied Donacia sagittariae in 1848, was 
unable to determine the function of the caudal spines and 
thought there was a delicate membrane, that is the spiracles at 
the base of the spines, stretched over the opening, and that 
there was an osmotic interchange of the air of the tracheal sys- 
tem with that on the exterior of the membrane; but, when 
we consider the size of the membrane and that there is water, 
not air, on its exterior surface, this suggestion is seen to be 
incorrect. Von Siebold, who studied Donacia linearis 
_In 1859, concluded that the openings at the base of the spines 
were functional spiracles, and that the larva obtained its air 
supply from the intercellular air spaces of the plant, and this 
was accomplished by the larva eating a hole into the tissue of 
the plant, into which it later inserted its caudal spines. From 
what follows it will be seen that von Siebold came nearest to 
the correct interpretation of the conditions existing here of 
any of the investigators. The next investigator to consider 
this question was Dr EK. Schmidt-Schwedt in 1887, who studied 
the larva of Donacia crassipes. He found that some 
larvae, kept in a breeding cage, would, when the cage was dark- 
ened, insert the tips of their caudal spines into the tissue of the 
roots, but removed them as soon as the cage was lighted again. 
