30 NEUROPTERA. 
litt.), Georgia ®, San Antonio [1 9], Texas® [3 ¢, 1 @ | (coll. A. NW. S.) *.—Mexico, 
Teapa in Tabasco (H. H. Smith: 2 ¢,1 9), Acapulco in Guerrero (A. Agassiz, 
M.C.Z.: 2.3); British Honpuras (Blancaneaur: 3 3); Guatemata, Livingston 
(Wilson, M. C. Z.: 1 3,2 2), Cahabon and Teleman in Vera Paz (Champion: 2 3); 
Nicaragua, Polvon in Dep’t Occidentale now Chinandega (McNeill, colls. McLachlan 
and M.C. Z.: 2 3). 
Since, up to the present time, H. tricolor has not been recorded from any locality 
farther south than Georgia or Texas, it is not surprising to find that the Mexican and 
Central-American individuals exhibit a number of differences from the most detailed 
description hitherto published, based on specimens from the United States. 
The size is smaller: abdomen, ¢g length 36-39 mm., ? 30-31; hind wing, ¢ 
25-26°5, 9 26-27. 
dg. The second lateral thoracic blackish band (7. ¢. that on the metepisternum) is 
sometimes broken into two narrow stripes, one above the other; the third band 
(2. ¢. that on the metepimeron) is, in some males, reduced to a small spot, or absent 
altogether : these do not appear to be geographical variations. 
In all the males the inferior appendages are one-third (not one-half) as long as the 
superiors, the brown border at the tips of the wings is very slight, and the pterostigma is 
much smaller, as when best developed it measures but one-half a millimetre in length 
and surmounts two cells. From this a series of gradual changes in the size of the 
pterostigma may be traced as in A. macropus, since some males have only half a cell 
beneath it, in others the pterostigma is little more than a slight clouding around a 
single more obliquely-placed postcubital (as in certain individuals of the “ variety ” 
heterosticta of H. macropus), while in others it is absolutely absent on some of the 
wings (Acapulco, Polvon). 
The basal red spot on the fore wings of the males is less rounded on its outer (distal) 
edge than in the few United States specimens which I have seen. The number of 
cross-veins may be fewer or more numerous than stated in the publications quoted. 
?. Compared with Selys’s description ? the sides of the thorax are differently coloured, 
and are very much as given by Kellicott!°, being a pale brownish-yellow, with a 
metallic green stripe on the lower half of the mesepimeron and, in most cases, a short 
metallic green streak on the upper part of the metepisternum. The sides of the eighth 
abdominal segment are not more distinctly yellowish than those of the other segments. 
Pterostigma present in all individuals examined, surmounting from somewhat less than 
one to two cells; colour white or obscure, pale brownish. 
Thanks to the kindness of Mr. James S. Hine, I have a female from Sugar Grove, 
Ohio, before me as I write, as well as others of the same sex from Texas (coll. A. N.S.). 
* If H. limbata, Selys, is to be regarded as a variety of H. tricolor, as I am inclined to treat it, we may add 
Illinois and Waco, Texas, to the distribution of the insect’. 
