i8o Memoirs of the Indian Museum. [Voi,. Ill, 



Size and Proportions. 



Arrow's statement (1907, p. 441) that '' A rather striking feature in which the 

 Passalidae differ from the Lucanidae, as from wood-feeding insects in general, is their 

 constancy of size " has proved not to be invariably true; and the fact that it does 

 appear tobe true of most species renders exceptions to the rule the more remarkable. 

 These exceptions are of special interest from many points of view and have been dis- 

 cussed as fully as possible below (pp. 262-5). It will be sufficient to point out here, that 

 since the very species which are most variable in size have also proved to be the most 

 variable in structure, the most abundant, and the most widely distributed species in 

 each of the several genera to which they belong, the fact of a great difference in size 

 between two specimens cannot be held to indicate that slight differences in unimportant 

 characters are likely to prove to have a definite taxonomic significance 



The actual length of a Passalid depends in some degree on the extent to which the 

 head is drawn into the prothorax, and the pro thorax over the mesothorax, but it 

 appears to be the most convenient indication of size that can be given. All my 

 measurements of total length have been taken from the front of the labrum (or, in the 

 genus Ceracupes, from the tip of the big horn) to the tips of the elytra. As the pro- 

 portion of length to breadth seems to vary greatly in almost all species, no breadth 

 measurements are given. 



Symmetry . 



Oriental PassaHds may be either symmetrical or asymmetrical. The degree of 

 asymmetry may vary slightly in different specimens of a single species ; it seems to 

 be an indication of the degree of specialization attained by different species or genera 

 rather than of the degree of their mutual affinity. 



Antennae. 

 The antennae are always ten-jointed. The basal joint is always long and thick. 

 Joints 2-4 are thinner and only about as long as broad ; the fourth occasionally bears 

 a rudimentary lamella. Joints 5-7 always bear at least the rudiments of lamellae, 

 but these are often so constructed that they appear to form only a continuous thicken- 

 ing of the shaft when the antenna is in the least degree curled (see text-figure i, A and 

 B), as it almost invariably is at this point in dried specimens. The three terminal 

 jomts always bear lamellae. The structure of the antennae of any one species is 

 usually constant within very narrow limits. The only exceptions to this that I have 

 met with are in the species A ceraius laevimargo the antennae of which are remark- 

 ably variable in form, and in a specimen of A ceraius grandis subsp. hirsutus in which 

 the fifth joint of the antennae is devoid of the well-developed lamella that it normally 

 bears in that species. The lengths of the different lamellae relative to one another are 

 used in distinguisting the species of Macrolinus one from another, and in separating 

 the genus Kaupiol^LS from the genus Labienus ; and I have used the same character 

 in defining the Gonatas group of genera. Beyond this I have rarely found either the 

 relative or actual lengths of the lamellae of anj' practical use ; for they do not seem 

 to be ahsohitely constant and they are difficult to compare in the case of insects which 



