243 



tween the two species. Those bred from the larvae sent by Mr. Barlow^ 

 together with the adults received from Mr. Beach, agree more closely 

 with H. varianus. The variation even in the structural characters of 

 the species is shown in that one female specimen in our collection has 

 in the right anterior wing four perfect submarginal cells and in the left 

 but three, the normal number, and in another female both anterior 

 wings have four submarginal cells; the other specimens are normal. 

 When it is remembered that the number of submarginal cells is used 

 to separate a group of genera, including Harpiphorus, Emphytus, and 

 Dolerus, the confusion likely to result from such variation may be bet- 

 ter understood. 



The male, of which we have but a single specimen, is much smaller 

 than the smallest female, and the sides of its flattened abdomen are 

 nearly parallel, differing markedly in this respect from the much broader 

 and pointed abdomen of the other sex. In size as well as color there 

 is a wide variation in our specimens, the length ranging from 10 mm to 

 15 mm . In color the flies are honey-yellow and reddish, with the thorax 

 and head more or less marked with black ; the former in typical speci- 

 mens being almost entirely black. The four terminal joints of the an- 

 tennae, the labrum, tegulae, scutel, feet, and portions of the legs, white. 

 The two basal joints and more or less of the third joint of the antennae 

 are reddish. The apical portion of the third joint in all of our speci- 

 mens and the fourth and fifth joints are brownish black. The basal half 

 of the wings is clear; the outer portion, smoky. 



This insect has been recorded from Connecticut, New York, Virginia, 

 Illinois, Massachusetts, and Canada; and E. testaceus, which is prob- 

 ably the same insect, from Pennsylvania and Virginia. To these lo- 

 calities we have added New Jersey, Missouri, and Kansas. In Cresson's 

 Catalogue of Described Hymenoptera it is accredited to the United 

 States and Canada. 



Of insect enemies, Mr. Jack mentions a number of Hemiptera ob- 

 served by him to feed on the larva, and he also observed, but failed to 

 secure, a minute fly which was apparently ovipositing upon its eggs. 



Certain species of Cornus (G. florida and G. mas) were found by Mr. 

 Jack not to be attacked by this insect, but the foliage of G. sericea, G. 

 alba, G. stolonifera, G. paniculata, G. sanguinea, G. asperifolia, and one or 

 two others were greedily devoured. Polygonum and Wild Grape may 

 be doubtfully added to the list of its food-plants. Our failure to get 

 the larvae received from Mr. Barlow to feed on these plants makes ad- 

 ditional observations on this point desirable. ' 



