;9o8] The Gains Corizits 135 



BASIS OF SEPARATION OF SPECIES. 



Many mistakes have been made in the past by placing too 

 much confidence in color characters in the determination of 

 species. This is very unsafe, as it would be difficult to find a genus 

 more variable and inconstant in this respect. In this paper 

 structural characters alone are used for this purpose. A careful 

 study of an abundance of material has shown that these are 

 reasonably constant. The form and position of the antenni- 

 ferous tubercles and rostral lobes are excellent characters to 

 separate some of the species. The nature of the transverse 

 suture on the pronotum, and the form of the scutellum serve for 

 others. The genitalia give us excellent characters for several 

 species, and the general form of each species varies but little, 

 though the size may vary, and in several cases is quite distinctive. 

 There is still another character that is good in a number of cases, 

 though it is difficult of access and hence little use has been made 

 of it. This is the nature of the dorsal sutures between the third 

 and fourth, and fourth and fifth segments. It is a character that 

 seems to be quite constant though at times the middle portions 

 of these sutures may be so obliterated that it is difficult to dis- 

 tinguish them. 



ARTIFICIAL KEY FOR THE DETERMINATION OF SPECIES. 



1. (a) Last segment of abdomen short and broad, trtincate in female 



and rounded in male hyalinus 



. (b) Not as above 2. 



2. (a) Antenniferous tubercles prominent; transverse suture terminating 



in a loop 3. 



(b) Not as above 4. 



3. (a) Insect large. Connexivum spotted crassicornis 



(b) Insect small. Connexivum unspotted viridicatus 



4. (a) Scutellum broad at the tip, rounded 5. 



(b) Scutellum narrow at tip and pointed (J. 



5. (a) Antenniferous tubercles broad, sternum not black scutatus 



(b) Antenniferous tubercles very long, reaching almost to apex of 



head tuberculatus 



(c) Sternum black indentatus 



6. (a) Connexivum unspotted . . . . ' lateralis 



(b) Connexivum spotted . 7. 



7. (a) Insect large, light colored validus 



(b) Insect medium, dark, last segment in female pointed 8. 



(c) Last segment in female not pointed 9. 



S. (a) Last segment very long pointed punctatus 



(b) Last segment medium nigristernum 



9. (a) Abdomen much broader than thorax, with black transverse band. . sidae 

 (b) Insect small, antennae short, scutellum large parvicornis 



