( 322 ) 



a point (PI. XVI. fit;. 4). The priiioipal function of the proventricle appears to me 

 to serve as a kind of sieve liy wliicii the food, which may have l)ecome luinped 

 together in the crop, is divided up into smaller particles. The proventricle cannot 

 cut up any hard substanci's. In one of the specimens of Arixenia a piece of chitiu 

 covered with long hairs was lying in the hind part of the proventricle. It was 

 unbroken, although its diameter nearly equalled the transverse diameter of the 

 proventricnlus. 



The stomach (sto, PI. XVIII. fig. 1) is not quite so wide as the crop. It is 

 asymmetrical at the base, bulging out towards the right side, forming a kind of 

 sack. The stomach of Forfiiuila was represented by Dufour* to be (|uite straight, 

 without any convolutions, and this statement is still being cojiied in te.xt-books. 

 In all the specimens of Vorjicula auricidaria, however, the apex of the stomach 

 forms one convolution together with the anterior portion of the small intestine, as 

 is the case also in Ilemimerus (PI. XVIII. fig. 2). As the crop of the oesophaijus 

 of Arixenia takes up so much room — its enormous capacity seems to indicate that 

 Arixenia takes a large quantity of food at a time and stores it, so to speak, in the 

 capacious oesophagus — the stomach is comjiletely coiled np so as to acquire but 

 little space. It forms nearly two convolutions, and a third is formed by the basal 

 portion of the small intestine. The coil is arranged like the convolutions of a shell, 

 the anterior half of the stomach forming the largest convolution, which is ventral 

 and gradually ascends dorsad. The centre of the coil is the most dorsal point of 

 the spiral. At this point the small intestine descends vertically, and then curves 

 backward, as indicated liy the dotted lines in our figure. The small intestine ends 

 in a large rectum, whose si.\ internal projections are long and narrow. 



The Malpighian tnbules (IZ/v, PI. XVIII. fig. 1) are very narrow. They 

 form a densely coiled np mass which lies on the top of the stomach and a similar 

 mass ])laced beneath the stomach. When the stomach is uncoiled the tubules are 

 found to be arranged in four bundles of about twenty tubules altogether. The 

 tubules of each bunch open in a very short common duct. The largest bunch is 

 inserted dorsally on the posterior (or left) side of the ape.x of the stomach, and 

 consists of ten tubules. A second bundle of five tubules is found subventrally on 

 the right side. There are further three tuliules placed on a short tube subdorsally 

 on the right side, and two similarly connected tubules subventrally on the left side. 

 The numbers vary very slightly in our specimens. The places where the dorsal 

 bundles are inserted are marked black in our figure. Many of the tnbules branch 

 off in twos and threes from a short common tube, as in the earwigs.f 



I have not found any salivary glands. 



Systematic Position. 



The agreement with the earwigs is too close to admit any doubt that Arixenia 

 is a kind of wingless Dermapteron. If the characters, however, which distinguish 

 liemimerus from the true earwigs are considered of sufficient weight for placing 

 Ilemimerus in a separate suborder of Dermaptera, we must erect a third suborder 

 for the reception of Arixenia. But I al)stain from giving a name to the suborder, 

 as there is still some uncertainty what name the whole order of earwigs should 

 bear. 1 have referred to them as Dermaptera, which name is the most commonly 



• I.e. 



t I find IG tuljes in Fnrficula aurieularia, arranged in 4 bundles (5, 3, 4. 4). The statements 

 by Dufour iind ticliiadler that there are about liO or 40 tubules are cerl.aialy erroneous. 



