REPORT ON THE PELAGIC HEMIPTERA. 33 



the middle of its length (PL III. fig. 22). These processes differ in shape in different 

 species, but deficiency of material to dissect has prevented me from describing them 

 minutely in more than two species (PI. III. figs. 23 and 24). It may be noted that in one 

 species the left horn is always bent outwards and forwards (PL III. fig. "22); in all the 

 other species examined by me both horns are symmetrical. The posterior angles of the 

 segment are in some species furnished with a tubercle-like prolongation; in others this 

 prolongation is situated between the angle and the base of the horn. 



The Third Segment. — Above, the visible part of this is very conspicuous as an 

 appendage to the end of the abdomen, in outline somewhat like the fluke of an anchor, 

 with the lateral angles more or less produced and the general shape varying according to 

 the species (PL III. figs. 20 and 25). The disk is longitudinally convexly elevated from the 

 base (which is sometimes slightly gibbous) to the obtusely conical apex. The sides are some- 

 what flat, or even slightly depressed. The apex of the segment, viewed from above or from 

 below, looks like an obtusely conical tubercle (PL III. figs. 25 and 26 ; and PL I. fig. 1, g a), 

 but closer examination shows that the lower half of it is a nearly circular very convex plate, 

 attached by the basal angles to the plate above. Between these plates is a horizontal 

 fissure (at the very apex of the abdomen), in which the intestinal canal opens. The plates 

 are therefore the podical plates. On dissection, the under surface of the upper or lozenge- 

 shaped plate shows a deep longitudinal hollow, in which the intestine lies (PL III. fig. 26). 

 The under surface of the extended sides is sometimes armed with short spines. 



The third seorment below is not continuous at the sides with the thiixl segment above, 

 and in its normal condition (that is, without having been dissected out) appears as a long, 

 oval, convex plate, overlapped at the sides and end by the upper plate, and additionally 

 protected at the sides by the horns of the second segment (PL I. fig. I, g a). In some 

 species the margin appears to be a little thickened, and to form a ledge on which the horns 

 rest. On dissection, the plate presents internally (PL III. fig. 27) a deep concavity, filled 

 by the horny case which contains the copulatory apparatus. The basal angles are 

 expanded and go uj)wards and inwards to meet the base of the lozenge-shaped plate 

 whose attachment is between them, the point of attachment being concealed by the pro- 

 duced hind margin of the second segment. 



The horny capsule (PL III. fig. 28) just referred to is, viewed from above, very convex, 

 rather strongly compressed laterally, and with a very polished surface. Dissected out and 

 viewed from the side, it is seen to be a semicircular case, with the sides somewhat flattened, 

 and the opening on the straighter margin of the semicircle. This straighter margin is not 

 exactly straight, but, so far as the chitinous portion of it is concerned, is for the basal 

 third straight, then widely concave, and then, for about the apical fourth, gently rounded. 

 Along the straight and concave portions of the margin are (in the dissection) the remains 

 of the membrane by which the capsule is attached, the apical fourth being free. It is 

 difficult to make out and to describe the structure of the apparatus contained in the 



(ZOOL. CHALL. EXP. — PART XIX. — 1883.) T 5 



