J. W. H. Harrison 207 



definitely. The case of Lymantria dispar and its subspecies japonica 

 produces evidence of the same value. 



Local races are in quite another category and no genus is more prone 

 to their development than Oporabia, the insects in nearly every habitat 

 displaying local facies. These races are clearly brought into being in 

 response to environmental forces of varying incidence which are not 

 powerful enough to push the insect on to an evolutionary plane palpably 

 higher than that of its relatives in other stations. The differences 

 exhibited may be in any feature — markings, habit, period, etc., each and 

 all being involved, — but they never attain to differences in structure 

 even of the low order existing between 0. autumnata and 0. filigram- 

 maria ; there is never any difficulty in assigning the insects to the species 

 to which they belong. So near do some genuine local races approach 

 what systematists pronounce the type that only very careful biometrical 

 study can demonstrate that difference in modal condition indicative of 

 the local race. 



In general, however, if a sufficiently great number of individuals be 

 assembled from any given area (restricted it may be in some cases but 

 of considerable extent in others) and placed alongside a similar as- 

 semblage from another point it will instantly be impressed upon one 

 that the two sets are not alike. Sometimes each individual in the one 

 differs from each in the other by some more or less tangible characters 

 or degree of development of those characters, and sometimes only the 

 majority of individuals show this difference ; nevertheless it exists, and 

 such diverse sets can only be regarded as local races. 



In Oporabia, whilst the differentiation of subspecies and local races 

 is fairly easy, great difficulty lies in the way of assigning any given 

 specimen in the absence of exact data to its correct group, so great is the 

 degree of fluctuating variation exhibited in wing markings. No species 

 or subspecies is free from it, although 0. christyi is least affected, 0. fili- 

 grammaria being next in order. But, let it be noted, even they can 

 cover the whole range. In all the ground colour can vary from almost 

 pure white to black; the markings in all may be absent, undecided, 

 scattered, condensed into bands, heavy or light and so on with every 

 possible combination of these and the ground colour ; even the dis- 

 coidal point achieves the apparently impossible by displaying great 

 variability. 



Haphazard as all this seems, it is in reality quite regular; indeed 

 so regular is it that all four forms possess parallel variations. So 

 obvious was this that when once I knew of the existence of the Carpet- 



