414 Mr. G. A. K. Marshall on 



" In the few specimens of Amauris echeria that I tried 

 I found that no juice was emitted, but they had a nauseous 

 taste and a strong smell, which reminded me somewhat of 

 that emitted by many Goccinellidx, But it was L. chry- 

 sippus which showed mo the futility of trying to arrive at 

 any definite conclusions from this line of research, for it 

 emits neither juice nor smell, and I could detect no trace 

 of any taste, unpalatable or otherwise, but the tissues have 

 a somewhat soapy feel to the tongue, which I noticed in 

 A. echeria and some of the Acrreas. The same may be 

 said of Mylotliris agatldna, though from its conspicuous 

 colouring, slow flight, and wide dispersal, I feel sure it is 

 an inedible species. 



^^ Malvern, Feh. 21, 1897. — Aera'a horta exudes a bitter 

 yellow juice from the thorax when it is injured, and this 

 juice permeates the costa of the fore-wing. The head and 

 abdomen do not appear to me to have any unpleasant 

 taste. Trimen refers to their smell, but my smelling 

 powers are not sufficiently acute to detect it. 



"Malvern, May 14, 1897, — Ahena amazoula feigns 

 death most persistently ; it has an unpleasant taste and 

 strongf smell not unlike that of the Coccinellidfe." 



28. Guy A. K. Marshall's Proof of Seasonal 

 Changes in Soutu African Butterflies of the 

 Genus Precis. (E. B. P.) 



A. Tntroduetion. 



The attempt will be made in the following section of 

 this memoir to explain these astonishing changes as due 

 to the adai)tation of a moderately distasteful and protected 

 genus in two directions — towards conspicuous warning 

 colours in the generations of the wet season, the time 

 when insect-food is abundant; towards procryptic conceal- 

 ment in the pressure and scarcity of the dry season. 



Facts which require for their interpretation the hypothe- 

 sis of adaptation in the direction of conspicuousness will 

 be brought forward, much use being made of the conclu- 

 sive proof only recently obtained by ]\Ir. Marshall, by 

 breeding the one from the other, that Precis simia is the 

 wet phase of P. antilo'pe. 



The distinct habits and stations of the two phases, their 

 relation to otlier seasonal forms of butterflies, the observed 



