OF THE WATER-BEETLE PELOBIUS TARDUS, 91 



food material. An examiuation of tlie .stomodteiini, however, 

 explained the mystery. 



The mouth-parts consist of a pair of sharply-pointed mandibles 

 which overlap when closed, the right ovei'lying the left. The 

 maxillte are palp-like in appearance, the maxilla itself forming 

 the basal part of the palp, and, when at rest, the whole structure 

 is contracted by being telescoped within the base (v. PI. II. 

 figs. 1 to 4). The labium has an ajjical piece, which, from its 

 appearance, I will call the "scoop " (s), and which, when at rest, 

 is contracted within the basal piece which bears the two short 

 labial palpi (v. PI. II. figs. 1 to 4). 



When the floor of the mouth is examined (v. PI. II. fig. 1), 

 it will be seen that at the base of the extensible part of the labium 

 is a small, projecting piece in the median line. This piece, which 

 I will call the " flap " (/), rises up in the " scoop," and when the 

 mouth is closed this flap stands up, and the lobes of the epi- 

 pharynx — this structure being split like a hair-lip — descend 

 from above and thus block the entrance. The anterior part 

 of the floor of the mouth is covered with minute projections 

 arranged more or less in rows radiating from a deep groove, 

 while the anterior part of the scoop is similarly studded. In the 

 posterior region of the floor of the mouth is a dark brown 

 chitinous plate (Ay/>) with a raised median piece posterior to it, 

 and I regard these structures as representing the hypopharynx. 

 On either side of the chitinous plate there projects inwards a 

 membranous lobe (mxl) which is extremely difficult to make 

 out ; I examined a large number of specimens before I could 

 satisfy myself that it really existed. Possibly these lobes 

 represent the maxilluljB. 



Immediately I^ehind the hypopharynx the mouth contracts into 

 the gullet, which is very short and opens out at once into the 

 cesophagus (v. PI. I. fig. 6). This again is very short, and there 

 is no pro-ventriculus clearly marked off at its posterior end, but 

 this region is armed with eight dark- coloured spines (v. PL II. 

 figs. 5 & 6), which project forward in the walls and posteriorly 

 are enlarged. The inner faces of these spines are armed with 

 strong, backwardly turned teeth which are specially numerous on 

 the enlarged bases, and evidently form the grinding-mill by menus 

 of which the food is broken up. 



With the exception of a pair on each side whose bases are 

 fused together, each spine is independent of the others. The 

 median ventral one is the longest, and extends forwards for about 

 two-thirds the length of the oesophagus, none of the other spines 

 extending more than half the length of the cjesophagus. 



To the o3sophageal walls, between the pro-ventricular spines, 

 are attached muscles which run to the head capsule, so that the 

 oesophagus is capable of considerable expansion and contraction. 

 In this case, therefore, almost the whole stomodjBum acts as a 

 sucking-pump, there being no pharynx marked ofi'for this purpose, 

 as is usually the case. 



