1S02.] no? 



as 4 and 5 put together; and the stigma is '.U times as long as wide, very 

 acutely pointed at both ends. Three or four specimens have a little white 

 pruiuesceuce still attached to their wings. Occurred very abundantly on 

 various forest trees in September. This is the largest known N. A. spe- 

 cies either of this or the following genus, P. pijrl expanding only .88 

 inch. That species is distinguished at once from vacjabunda by its wing- 

 veins being all black. 



PEMPHKIUS Hartig. (Fig. G.) 



Pertiplinjun pt/rl Fitch, (apple-tree roots) N. Y. Hep. T, p. 0. 1 



species. 



The species described below under this genus differ from Pcmjihigufi 

 as limited by Koch, chiefly in the stigma being short or rather short, not 

 "narrow and long." I do not consider this character of much generic value 

 in Apliklsc. In Aphis avenn: the stigma is four times as long as wide; in A . 

 mali only twice as long as wide; and the two extremes of length in that 

 genus seem to be connected by an unbroken series of intermediate grades 

 as is partially exemplified in the few species described above. 



I suspect that all the species properly referable to this genus live under 

 arround and derive their nourishment from roots. Authors have lonji no- 

 ticed that Aphidian insects are found in ants' nests, and Westwood states 

 that all species found in such situations are apterous. (lutrod. II, p. 441.) 

 I have succeeded in breeding to the winged state one species found in 

 the nest of a common yellow ant, described below as Formica aphldicola. 

 and I have found numerous winged specimens of another species on vari- 

 ious occasions in the nests of the same ant, in company with prodigioixs 

 numbers of larvae. Both species appertain to Pemphigus, with the ex- 

 ception of the above noticed differences in the stigma. I have also ascer- 

 tained from repeated observations the very curious fact, that the ants fetch 

 the larvae of Pemjjhigus formiccforum mihi, home to their nests, from the 

 roots on which they feed, and place them in little clusters of 50 or 60 in- 

 dividuals, where they soon elaborate such a dense mass of white cottony 

 matter as to entirely conceal them. The proof of this rests upon the cir- 

 cumstance that I have often noticed clusters of these larvae— some cover- 

 ed with flocculent matter, some naked — in nests located in honey-combed 

 stumps more than a foot from the ground, where there are no roots for 

 them to feed on. They are also found on the inferior surface of flat stones 

 covering the nest; and in both cases they are generally jdaced close to 



