128 THE STUDY OF INSECTS. 



J. Head with a transverse incision in front of the 

 ocelli, which are always present. (Fig. 17$.) 



p. 143 BERYTIDiE. 



JJ. Head without transverse incision. 



K. Membrane with four or five simple veins 

 arising from the base of the membrane; the 

 two inner ones sometimes joined to a cell near 



the base. (Fig. 146.) p. 142 Lyg^eid^e. 



KK. Membrane with many, usually forked 

 veins, springing from a transverse basal vein. 



(Fig. 147.) p. 143 Coreid^e. 



DD. Antennae five-jointed. 

 E. Scutellum nearly flat, narrowed behind. 



F. Tibiae unarmed or furnished with very short spines. 



p 144 Pentatomid.e. 



FF. Tibiae armed with strong spines in rows. p. 145. 



CYDNID.E. 



EE. Scutellum very convex, covering nearly the whole ab- 

 domen. 



F. Lateral margins of the scutellum with a furrow in 

 which the edge of the wing-cover fits when closed. 



p. 146 CORIMEL^NID^E. 



FF. Lateral margins of the scutellum without a furrow. 



p. 146 SCUTELLERID^E. 



In the following pages we have discussed the families of the HeLe- 

 roptera in the order in which they are commonly arranged by writers 

 on these insects. The aquatic families are placed first, the semi- 

 aquatic next, and the terrestrial last. We do not believe, however, 

 that this represents well the lines of development of these insects. It 

 seems probable to us that some of the terrestrial forms, as the Penta- 

 tomidae, resemble the ancient Heteroptera more closely than do the 

 aquatic forms ; and that the aquatic forms are really very highly spe- 

 cialized and represent the summit of one of the lines of development. 

 Perhaps the Scutelleridae represent the summit of another line. 

 But as we have been unable to give the matter sufficiently thorough 

 study to warrant our proposing a new arrangement, it seems best to 

 follow the old one, merely indicating our belief in the necessity for a 

 revision of the suborder. 



