90 the entomologist's record. 



I am about to deal with. The body is covered with long, thick hair to 

 the same extent as in the former and as described by Linneus in 

 Fauna Suecica ; the dark scaling is usually more indigo in hue and not 

 as glossy as is usually the case in the races described below ; the red 

 is not of so clear and bright a carmine, and on the underside of the 

 forewings it does not form as thick and uniform a red patch, so that 

 the shiny surface of the chitin is seen more, through the thinner 

 scaling. The Cancale examples are even larger than the average 

 filipendulae, but others I have from the Isle of Wight are smaller. 



A race quite similar to the one just described, so that it can quite 

 well be called nymotypical of filipendulae too, is that which is found in 

 most Alpine localities and which constitutes a transition from the 

 extreme mannii and paidula of the highest altitudes to the races of the 

 lowest and warmest valleys, where pulchrior is found N. of the water- 

 shed and microchsenheimeri or ochsenheimeri are S. of it, in Italy. Local 

 variations, observable on comparing series of this race filipendulae from 

 different regions, can be designated as trans, ad arctica, or mannii, or 

 paulula, or pulchrior. 



Race pulchrior, mihi. It is quite surprising that no writer 

 should as yet have noticed the sharp difference between the northern 

 filipendulae and the race, Avhich is generally distributed in Central 

 Europe, especially in the lowlands. Its characteristics are the absence 

 of a frontal tuft, its thorax and abdomen devoid of hair and covered 

 with scaling as glossy as those on the wing, the brilliant gloss of the 

 dark scaling of the latter, the somewhat clearer, more vivid and warmer 

 tone of red, the denser scaling, particularly noticeable on the underside 

 of the forewings. Specimens collected by C. Hofer in the neighbour- 

 hood of Vienna (Klosterneuburg), in July, are good representatives of 

 this grade in the variation of filipendulae and I select them as "types." 

 Many entomologists have been struck by the notable variation in the 

 size of this insect. The typical size may be said to be 30-81mm. in 

 the male and 34-36 in the female. In some localities 29 and 30 is the 

 prevalent size and these specimens look a great deal smaller than this 

 difference seems to show, because of their frail build. I have some 

 from the Gumpoldskirche (a hill near Vienna). In many cases one 

 might well talk of a subrace paupercula, mihi. In mountains series 

 of specimens are to be collected which can be described as trans, ad 

 filipendulae and the same forms are prevalent towards the north, 

 although marked differences occur in localities even not far from each 

 other. Towards the south, on the contrary, as might have been 

 expected, races describable as trans, ad pnlcherrima make an appearance 

 and these then gradually merge with stoechadis, completing the series. 



Race pulcherrima, mihi. Oberthur has already noted (I.e., 

 p. 500) that the most brilliant French race of filipendulae is that of 

 Dompierre-sur-Mer (Charente-Inferieure), in the S.W. of France. I 

 have a large series from this locality, collected by P. Boule in the first 

 half of June and I think no other race of subspscies filipendulae could 

 be larger and more dazzling in colouring, so that I choose it as typical 

 of the last grade, before stoechadis. Males vary from 31 to 35mm., 

 females from 36 to 40mm., in expanse. The red spots are always very 

 large, often confluent in pairs and even the usually extremely rare 

 aberration in which they are all united by a broad band along the 

 cubital nervure (ab. confluens, Obth.) is comparatively frequent. The 



