1916] Aquatic Hemiptera 357 



Reverting now to the respiratory tube in this family of water- 

 bugs, this, as is well-known, consists of two parts fringed with 

 hairs, and grooved in the inner surface, forming thus two half 

 tubes longitudinally divided. When the insect is submerged 

 these two half tubes are kept together by the pressure of the 

 water, the hairs serving to prevent leakage of air at the junction 

 or seam between the two halves of the tube. These two halves 

 are nothing but the lengthened peritreme of the seventh pair 

 of abdominal spiracles, and they vary in length with the genus 

 and species and even with the individual. It has been experi- 

 mentally demonstrated that the shortening of the tube in any 

 individual in no wise interferes with its respiration. But what 

 does not appear to be so well known, if known at all, is that 

 the halves may be of unequal length without at all affecting 

 the usefulness of the tube. I have observed this in a specimen 

 of Ranatra thus maimed, which was kept alive in an aquarium 

 for a considerable period and which suffered no inconvenience 

 from the inequality of the halves of the air tube. As the 

 bug hangs head down from the water plants, the apex of 

 the tube pierces the surface film and leads air in to the round 

 spiracles situated at the base of each filament, from which arise 

 the two main longitudinal tracheal trunks. In Nepa, the tube 

 is short, as it is in Cercotmetus, a near relative to Ranatra, 

 and in Curicta, the form that bridges the gap between Ranatra 

 and its allies and the Nepa-\ik.e forms; on the other hand, in 

 Ranatra it is long (with perhaps one or two exceptions) , as well 

 as in Laccotrephes and other exotic genera. Here is a distinct 

 adaptation to a preferred habitat, as Ranatra and the other long- 

 tubed genera are deep-water dwellers, while Nepa and its short- 

 tubed congeners is to be found in shallows among grasses. 

 In the Nepidce as a whole there is a most peculiar respiratory 

 structure — the so-called false spiracles, which are six in number 

 and are situated in the connexivum at the third, fourth and 

 fifth abdominal segments. These false spiracles are large and 

 oval in outline, and quite visible to the naked eye. Under the 

 magnifier they are seen to consist of an opaque plate set in a raised 

 margin and provided with round perforations which, in turn, 

 are occluded by a diaphanous membrane. The true spiracles 

 in these segments of the abdomen, as well as in the second 

 segment, are nonfunctional, the former being set in the widened 

 peritreme of these colander-like openings. No experimental 



