28 



This butterfly is greatly influenced in the time of its emergence 

 according to seasons. In the first week of May, 1888, I found it 

 abundantly and in fine condition at Hyeres, whereas in 1894, an early 

 season, by May 6th not one was to be seen, and I was almost too late 

 for the larvae, which were then full-fed. I am told the butterfly was 

 out this year on March 15th at Hyeres. 



Var. cassandra is not at all uncommon at Hyeres, but it is extremely 

 local. I once came upon a colony in a swampy piece of ground near 

 the river Gapeau, on the north side of Hyeres, where it was very 

 abundant. I imagine wherever the food-plant, Aristolochia rotunda, 

 occurs plentifully the butterfly is to be found. This form of T. 

 polyxena occurred in some numbers in a friend's garden on the out- 

 skirts of Hyeres, the food-plant being common in the vicinity. It 

 was here I found the larvae from which the series now exhibited was 

 bred. My friend Mr. Raine, who has frequently reared cassandra, 

 tells me that he had never bred any ichneumons from the pupae ; my 

 experience, however, was the reverse, for I bred more ichneumons 

 than butterflies. Var. ochracea is not at all scarce, and is merely a 

 form in which the ochreous ground-colour is darker. 



Thais rumina, the third and last species, occurs in Southern Spain 

 and Portugal, and is not uncommon at Gibraltar. North of the 

 Pyrenees, along the Riviera and as far north as Digne, in the Basses 

 Alps, we find the variety medesicaste. The markings in this form and 

 the type are practically identical, the difference appearing to be 

 merely a question of intensity of colour, the ground-colour in the 

 type being deeper ochreous, and the red blotches much brighter. 



Medesicaste was fairly common at Hyeres in the beginning of May, 

 1890. This year three specimens were taken there on March 17th. 

 At Digne (which is about 2000 feet above the sea level), probably its 

 northern limit, the species was not uncommon during the first week 

 in June, 1890. At the same time I found larvae in all stages on their 

 food-plant, Aristolochia pistalochia, the young ones on the under sides 

 of the leaves and in the tubes of the curious flowers, the full-grown 

 larvas lying concealed at the roots. It is rather a curious fact that 

 bred specimens of medesicaste are so much larger and finer than 

 captured ones. I can only account for the circumstance in this way : 

 the food-plant grows on hill-sides much exposed to the sun, and in 

 dry seasons probably becomes parched up, and the larvae, being 

 sluggish, suffer in a corresponding degree ; whereas by rearing the 

 larvae one collects all the succulent plants one can find in shady 

 places. I visited Digne again in 1894, but medesicaste was extremely 

 scarce. Local collectors told me that they believed the species had 

 suffered from the effects of the drought of the previous summer. 



Var. canteneri of Staudinger is a variety occurring in Spain and 

 Northern Africa, and differs from the type in having the ground-colour 

 very dark ochre. Var. honoratii is a .very beautiful variety, in which 

 the red blotches are confluent. It is only recorded from Digne, and 

 seems to be somewhat rare, to judge from the high prices local collec- 



