358 THE CANADIAN ENTOMOLOGIST. 



Tliis primitive form would, as an adult, be rather strongly and heavily 



built ; the tegmina would be sharply divided into three areas, viz.: clavus, 



corium and membrane, the latter being membranous and numerously 



veined, the two former coriaceous. The antennae would be 5-segmentate,t 



the base concealed under the head ; the labium {rostrum) would be free, 



and would be composed of four segments ; the tarsi would have three 



segments, and the metasternal glands would be very strongly developed. 



The nymphs would also possess well-developed stink-glands, three in 



number at least, and paired. 



This development of the stink-glands, the sharp limiting of the mem- 

 brane from the coriaceous part of the tegmina, the quadrisegmentate 

 labium, etc., are, in my opinion, fundamental characters of a typical 

 Heteropteron, and it is in those forms which are obviously highly modified, 

 that any departure from the above is specially shown. Entirely mem- 

 branous tegmina are found in Gerridse and Enicocephalidse, obviously 

 highly-modified families. It is in the Gerridse also, and the true water- 

 bugs, that degeneration of the metasternal stink-glands has proceeded 

 so far. 



Osborn, in his classification, has allied the Miridse with the Pyrrho- 

 coridse, the Acanthiidse with the Gerridse, and the Aradidae with the 

 Clinocoridpe, but these alliances are certainly superficial, not phylo- 

 genetic. 



The classification adopted by Distant is entirely artificial, as well as 

 inaccurate in details. He retains the discredited divisions of "Gymno- 

 cerata" and ''Cryptocerata," and proceeds at first in the former by 

 separating off the Gerridae and the Naeogeinaj, because the sternites are 

 clothed beneath with silvery-velvety pubescence ! The Cimicidce are then 

 separated off by the scutellum reaching at least to the base of the mem- 

 brane (though this is not the case in at least some Urostylinre), the 

 scutellum in the other families being stated not to reach the base of the 

 membrane, although this is certainly the case in most, if not all, 

 Macrocephalidae and Aradidae. The Acanthiidte are associated with the 

 Reduviidae, the Aradidae with the Macrocephalidae, and the Nepidae with 

 the rest of the Waterbugs. 



tXhe possession of only four segments in almost all Heteropterous nymphs 

 may prove that the adults of my supposed primitive Heteropteron really had only 

 the same number, but it is a matter of little consequence. 



