4 LUCANID^ AND PASSALID.E. 



trochanter, and the inner side of that joint usually bears a 

 hard straight ridge extending from its base to the tip of this 

 projection. Highly magnified, this ridge is found to be 

 transversely broken up or milled, like the edge of a com. 

 Tlie leg is usually directed a little forward, so that the ridge, 

 wliich is used like a violin bow, rests upon the base of the leg 

 in front of it. The coxa of that leg is crowded with hard 

 granules, so that it may be compared to sand-paper, and the 

 effect of drawing the sharp transverse blades of the " bow " 

 across this jjlate is to set up vibrations which produce a 

 fairly high-pitched note. The granules upon the middle 

 coxa have a definite arrangement which varies according to 

 the species. The structiu-e found in the European representa- 

 tives of the genera Lucanus and Dorcus are shown in the 

 beautiful plates of the Danish work on beetle larvae (Scliiodte — 

 De Metamorphosi Eleutheratorum Observationes). In 

 Lncamis the outer edge of the granular area is formed by a 

 single row of larger elevations placed side by side. In Dorcus 

 the arrangement does not differ greatly, but certain other forms 

 less nearly related show well-marked differences in the dis- 

 tribution of the granules. In some these form rows instead 

 of being distributed over the whole surface. In a recent 

 paper (Stett. Ent. Zeit., vol, xcvi, 1935 p. 178) Dr. van Emden 

 has described the distinctive features of the larvae of a number 

 of different genera and Mr. J. C. M. Gardner (Indian Forest 

 Records, vol. i, 1935, p. 6) has compared various Indian 

 larvae identified as belonging to Dorcus, Hemisodorcus, Proso- 

 pocoelus and Euryimchelus, which he says " might all belong 

 to one genus," thus confirming the view taken in the present 

 work. A few other Indian larvae have been described by 

 Dr. Gravely (Records of the Indian Museum, vol. xii, 1916, 

 p. 137). In the genera Jigus and Xigidius the stridulatory 

 surface is less well developed and it seems doubtful if the 

 apparatus is actually functional. 



Passalid larvae also are rather smooth, the body-segments 

 being without the transverse folds found in most Lamellicorn 

 larvae, but, in maikcd contrast, not only with the Lucanid^ 

 but with all other known Lanu^llicorns, they are active creatures, 

 able to crawl from place to place. Tlio body is com2>;n*atively 

 long and straight, the ten abdominal segments are alike, tlie 

 jiosterior ones not enlarged, antl the ti>rminal one is witliout 

 the usual spiny " raster " on its \n\\vr surface. The anal 

 opening has the normal transverse direction, unlike that of 

 LucANiD^. The most remarkable characteristic is in the legs, 

 which appear to be only four in number. A very close 

 examination is necessary for the detection of the tiny hind 

 legs, seemingly rudimentary, wliich do not jiroject downwards 

 but lie close to the sides, where tliey extend only as far as 

 tlie bases of the second pair. When magnified these miinito 



