31(S Dr. T. A. Cliapiuau 07i 



approach to the hind margin, no specimens I have seen of 

 either species are at all near to (jrandis in this respect. 

 Gran(U>i agrees with iojtro(I((cf.ijli's in the inner dark half of 

 the fringe (continuous in f/raj^hodadylns), being broken up 

 into spots. 



On examining the ancillary appendages for some indi- 

 cation as to the relationship of these forms, one is met by 

 the remarkable fact, that there is very little difference in 

 these structures in the several species. The clasps are 

 very remarkable and comjjlex structures, but except in 

 some slight ditfcrences in proportions and curvatures are 

 quite alike not only in these gentian-feeding species, 

 but species of other groups such as p^crodadylus and 

 hipundidactylus are practically the same. 



I find, however, certain differences in the tegumen that 

 appear to be constant. These will be best understood 

 by reference to the plates (XV) ; it will be seen that 

 the tegumen of grandis has little affinity with that of 

 coprodadyl'u?i, and is much more of the type of that of 

 fjra] )]( odadylus. 



There is another difference that is of considerable im- 

 portance, viz, the great size of these appendages in grandis, 

 as compared with the other two species. (Compare Plates 

 XVI and XVII, where all are equally magnified.) Were 

 grandis merely an aberrational form, the large size of the 

 imago generally would be accompanied with appendages 

 of normal size. It is at least the rule for the appendages 

 to preserve a normal development, notwithstanding great 

 variation in size otherwise, so far as my observations go. 

 This is true of specimens from the same region, but not 

 ^always of geographical varieties, as of Swiss and British 

 Erdna ivthirqis. 



It is certain that though coprodadylvs occurs on the 

 same ground with grandis, it is physically impossible for 

 them to be syngamic, nor as a matter of fact were any 

 intermediate specimens taken. These specimens oi grandis 

 were taken July 21) th, 1905, on the top of one of the spurs 

 of the hill opposite Larche, Basses Alpes, at a heiglit of 

 rather over 6,000 feet. Some were taken on the wing, 

 others were bred from pupa? found together with empty 

 cases on the leaves of Gentiana ln/ca. It was remarkable 

 that no trace of where the larva; had fed on the plants could 

 be found ; the plants were almost all the large clumps of 

 root leaves, without flowering stems; though a few had 



