STUDY OF VARIATION IN THE RACES OF ZYGAENA FILIPKNDULAE, L. 87 



naming them bavarica, one does not quite understand ; his experiment 

 is the very proof that those insects belonged to one only species. I 

 think we can now come to a conclusion which will clear up many of 

 the mysteries of the past and which will even satisfy Oberfchiir's 

 creationist's convictions : different species can produce individuals, to all 

 external appearance, exactly similar to each other. Experts of the 

 various genera will, of coarse, reduce specimens of this sort to an 

 extremely small percentage, but the principle seems a sound one to 

 start from ; admittance of one's incapacity is a good step towards real 

 knowledge. A few years ago such a conclusion might have seemed a 

 wild hypothesis, but the study of the genitalia, especially in the 

 Hesperia, has proved that specific difference can exist without revealing 

 itself by external features in all the individuals. It may be dis- 

 heartening, but if we give up the practice of naming single specimens 

 and we base our judgment on series of them collected together, 

 hopeless cases will be limited to a very small number of individuals 

 from localities where two allied species fly together ; the vast majority 

 are, of course, recognisable. 



Starting from this point of view Oberthiir need not conclude, as he 

 does at pages 521, 549 and 550, that ' ; at Vemet-les-Bains every grade 

 of transition exists, in the most insensible and incontestable way, 

 between dubia and lonicerae and at Cauterets between lonicerae and 

 filipendulae.'" Judging from the experience I have gone through with 

 the Italian races, I think he will end by finding, like myself, that the 

 line can be drawn between the species or perhaps that one or two 

 specimens of his series will remain doubtful to the naturalist's eye, 

 although there is little doubt that their offspring, had they reproduced, 

 would have belonged distinctly to one or the other species, just as much 

 as that of any of the most distinct specimens of the series. 



As to anceps, of which I have a series before me, the case is quite 

 different : both Querci and I have perfectly satisfied ourselves that it 

 contains no loniierae and no trifolii and that it is filipendulae, which 

 varies to an extent very unusual in this subspecies and produces 

 frequently some lovely loniceraeformis individuals and others which 

 might be called trifolii for mis on account of their shorter, broader wings 

 and antennas with a more stunted point beyond the club. This 

 frequency of five-spotted individuals seems very natural, considering 

 the proximity of the locality of anceps to the region where filipendulae 

 is replaced by most highly characterised stoechadis. 



By these general observations on filipendulae and its near ally 

 lonicerae I hope I have taken one step towards showing that the 

 Zygaena are not as abnormal, compared to other genera, as they had 

 till now appeared. In papers on other species I will try and clear up 

 more problems about them, and in a general paper on the genus I will 

 try to show what an inspiring one it is, when their variations are com- 

 pared, and how useful it is to make out the few simple lines they 

 follow, because it enlightens one on the more complex phenomena 

 found in other Lepidoptera. I will now try and summarise the 

 geographical variations of filipendulae. 



Before doing this I must however explain the method I suggest 

 adopting to designate the gradations of geographical 

 variation. We are beginning to have quite a tolerably good 

 knowledge of the various aspects, that many species acquire in the 



