332 THE PLACE OF MIMICRY 



ninae extensively mimic each other, and are extensively 

 mimicked by other Lepidoptera. The Heliconinae which 

 resemble the Ithomiinae belong to the same group as 

 those which resemble other Heliconinae. The likeness 

 is even closer in some pairs made up of Heliconines 

 than in any made up of an Ithomiine and a Heliconine. 

 There is no good example of Procryptic colouring in the 

 whole Sub-Family, and the Heliconine types of Apose- 

 matic colouring are far more glaring than those of the 

 Ithomiinae. When, however, we inquire into the relative 

 numbers, it is seen at once that the Melinaea as an adver- 

 tisement is hundreds of times as efficient as the Heliconius, 

 because it is hundreds of times as numerous in the perfect 

 state. With Mr. Kaye's kind help I obtained two days' 

 captures of butterflies from the locality near the Potaro 

 River, British Guiana, where this wonderful Mullerian 

 combination has been chiefly studied. It was arranged 

 that butterflies were to be taken as they came, without 

 any selection. On the first day, August 28, 1903, just 

 323 butterflies were captured, of which 253 were Melinaea 

 mneme, while 2 were Heliconinae — 1 Heliconius vetustus, 

 and 1 Eueides nigrofulva. Of these, the first-named very 

 perfectly resembles the dark-hind-winged forms of the 

 Melinaea, while the Eueides is a far more outlying member 

 of the combination. On the second day, February 23, 

 1 904, out of a total of 325 butterflies, 220 Melinaea mneme 

 were taken, but not a single Heliconine. 1 Mr. Kaye's 

 experience, extending over many years, quite confirms 

 the extraordinary difference in numbers which was shown 

 by the results of these two days' captures. It may be 

 argued that the very numbers of the Melinaea imply a 

 higher degree of unpalatability, but this is by no means 

 necessarily the case. Relative numbers are determined 

 by many other causes, such as fecundity, attacks (chiefly 

 parasitic) in the earlier stages, &c. 



We are led to conclude that during the period of gradual 



approach the varieties of H. numata which were most 



conspicuously different from the pattern of Melinaea 



mneme suffered most, upon the whole, from experimental 



1 Proc. Enl. Soc, Lond., 1906, p. lxi. 



