22 LUCANIDJE AND PASSALID^. 



those of C. carinalus. In other cases the isolated phase 

 seems to be very rare. The corresponding phase of the 

 Phihppine C. alces is represented in the British Museum only 

 b}^ a single specimen captured nearly a hundred years ago, 

 and I am not aware that a second has ever reached Britain. 

 The Indian Dorcus spencei (Plate IX, figs. 5 and 6) is another 

 species of which, during very many years, this phase has only 

 once been fomid. Another Indian insect, closely related to 

 D. spencei, is of rather particular interest. This is D. poly- 

 morphus, which is abundant in the Darjeehng district, from 

 which I have seen about 80 males, all but three of them 

 belonging to the variable phase (Plate II, fig. 7). Small 

 specimens have flat triangular mandibles, the inner edges 

 of which are straight and can be brought together from base 

 to apex. In larger specimens they are separated near the 

 base but in the anterior half remain capable of close contact. 

 In two of the 80 specimens the mandibles have an entirely 

 diflferent form, fig. 5. They are slender, curved and far 

 apart, so that only the tips can be brought together, and 

 their inner edges bear only a few scattered teeth, instead of 

 the close rank found in the other phase. There is also an erect 

 tooth upon the upper surface, of which no trace aj)pears in 

 the ordinary form. The remaining specimen (fig. 6), in the 

 Oberthur collection, is a remarkable one. Like the two 

 just mentioned, it is of the maximum size. The left mandible 

 is in every respect that of the rare isolated phase, while tliat 

 on the right is identical with that of a similar-sized examj^le 

 of the ordinary phase. 



I have seen only one other Lucanid which, hke the last, 

 combines in itself both the constant and inconstant jahases. 

 This is a male in the British Museum of Dorcus forceps, an 

 insect mhabiting Borneo and Sumatra. In this case the 

 right, instead of the left, mandible is that of the isolated phase 

 and the left is that of the variable phase. I have learned from 

 the late M. Oberthur that in his collection is a male Dorcus 

 suturalis in wliich is combined the two forms of mandible I 

 have described above. 



Dimorphism of this peculiar kind is confined to no particular 

 region. In Madagascar Dorcus serricornis, a species related 

 to D. polymorphus and D. forceps, has two similar male phases, 

 but of twenty-six male specimens only one represents the 

 isolated jihase. 



Although the predommance of the inconstant form is the 

 general rule it is not invariable. The West African Dorcus 

 faber is an interesting excejation. I have seen 16 male examples 

 of this, 13 of which have long, slender, strongly curved man- 

 -dibles, meeting only at their tips, while only tlu-ee have the 



