892 DR. F. LEUTHNER ON THE ODONTOLABINI. 
But an entomologist who collected a long series of Cladognathus or Odontolabis in 
their native haunts would arrive at very different results. If collecting in Further 
India or the Sunda Islands, for instance, he would find such great variations in the 
mandibles of the males that he would be inclined at first, like the older authors, to 
regard them as belonging to several distinct species, till some fortunate accident 
enabled him to obtain a complete transitional series from one extreme form to the 
other, and he discovered that the same variations occurred in several allied species. 
He would next discover that the middle-sized forms have generally stronger and 
thicker mandibles than the largest and smallest, and that these species could not be 
considered simply variable, like the Lucanini and Dorcini, but must be regarded as 
polymorphic *. 
Thus Neolucanus castanopterus® constantly exhibits only one form of mandibles 
(Pl. LXXXIV. fig. 13), WV. saundersi has two forms (Pl. LXXXV. figs. 13, 16), Odonto- 
labis siva has three forms (Pl. LXXXVI. figs. 1, 3, 6), which are much more 
distinct in O. brookeanus (Pl. XCV. figs. 13-15), while O. alces exhibits four forms 
(Pl. LXXXIX. figs. 1, 3, 5, 6). 
As already mentioned, these stages are completely connected by transitional forms in 
a number of species (for instance, in 0. alces and Heterochthes andamanensis). But 
in other species this never occurs. In O. brookeanus no transitional forms between A 
(Pl. XCV. fig. 13) and B (fig. 14) have ever been observed. Nor do we ever meet 
with transitional forms bridging over the gap between the telodont and mesodont forms 
of O. sinensis, O. cuvera, &c. 
(3) The Question of Species in the Lucanide. 
Whoever studies the entire group of the Lucanide, whether from a systematic point 
of view or from that of comparative morphology, should understand that we have here 
to deal with a plasticity of material which does not allow of our discovering such 
definitely fixed specific characters as we are accustomed to look for in other groups. 
Wherever we seek for such characters we find them inconstant. It follows that in 
this family the definition of a species is more extensive than in other groups of insects, 
and that the descriptions must be differently arranged if they are to be of any perma- 
nent value. The old authors contented themselves with drawing up a description of a 
single specimen. ‘This renders it extremely difficult, if not impossible, to identify many 
of their species with any certainty, as those only can fully realize who have attempted 
to work at historical entomology. The study of types is more important in the 
Lucanide than in almost any other group, for without this assistance many riddles 
would remain insoluble. 
Asasingle specimen gives us a very imperfect idea of a species, we must describe 
1 Vide Parry, Trans. Ent. Soc. Lond. (8) ii. p. 66 (note). 
2 Also WN, laticollis (Pl. LXXXIV. fig. 1) and Odontolabis latipennis (Pl. XCVI. fig. 1). 
