ARGIA. | 69. 
stripe is fused with the humeral stripe proper at its lower, but not its upper, end. 
The colour of the labrum appears to vary but little; when no statement is made 
concerning it either in the Key to the Species or in the specific description, it is to be 
understood as being pale. | 
The measurements of the pterostigmata have been made with an eye-piece micrometer 
in a low-power compound microscope. ‘The dimension given is the length along the 
costal margin. It is not pretended that the extremes of size have been given for each 
species, but merely the results obtained from a few individuals taken at random. 
Even so they will probably be found useful, as the size of the pterostigma is, toa 
certain degree, a specific character. 
Much time and care has been devoted to the gathering of statistics relative to the 
number of antenodal cells and of those surmounted by the pterostigma. Nearly every 
specimen enumerated under each species has been tabulated in regard to these points, 
involving an examination of about five thousand wings. These two details of neura- 
tion were employed by Hagen? as means for subdividing the genus into sections. 
It will be seen that considerable variation in these numbers exists. On the whole, 
antenodal cells are less variable than subpterostigmal cells, but neither feature is 
sufficiently constant for the purpose for which Hagen used them, and the same negative 
result is reached on testing a third of his characters, viz. the place of termination of the 
superior sector of the triangle. All neurational differences observable among various 
species of Argia are too variable to serve as a basis for subdivision of the genus. 
After many efforts to arrange the species of the present fauna into natural groups, 
the attempt has been abandoned. ‘The alliances suggested by one set of characteristics 
appear to be contradicted by others having equal weight. The key here offered has 
very little pretension to expressing natural relationships. It is only given in the 
belief that it is better than no key at all, and as being the least unsatisfactory of those 
which I have successively drawn up and discarded. | 
Artificial Key for the identification of Mexican and Central-American 
Species of Argia *. 
Males. 
(N.B.—AlIl identifications made by this key should be checke.] by comparison of the appendages 
with figg. 27-62, Tab. 1V.) 
§I. Total area of dark colours on abdominal segments 3-6 and thoracic dorsum 
greater than the total pale area on the same parts. 
1. Thoracic dorsum not brilliantly metallic. 
* The species in square brackets marked with an asterisk have been taken in Texas or Arizona, and may 
occur in Northern Mexico, but have not yet been observed within the faunal limits of this work, 
