90 SOUTII-AFIUCAX BUTTERFLIES. 



Arabia, to which four of them extend ; while three others are common 

 to Africa, Arabia, and India. One section {Etrida gronp) is peculiar 

 to India, and one (MancmJiari group) to Madagascar ; while another 

 (Uvcmthe group) is common and peculiar to those two countries. India 

 possesses representatives of six sections and Madagascar of four. Of 

 the ninety species here enumerated, fifty-five appear to be peculiar to 

 Continental Africa, or sixty-nine to the entire Ethiopian Region, includ- 

 ing Aral^ia and Madagascar, which latter has yielded four endemic 

 forms, while nine appear to be limited to Arabia. In South-Africa I 

 recognise twenty-nine species, of which five only are not recorded as 

 occurring beyond the tropical limit. 



In these statements it must be noted that the numbers can only bo 

 regarded as generally indicative of the actual distribution. There is, 

 perhaps, no genus of Ijutterflies more puzyAing to deal w^ith than Tcra- 

 colus, owing mainly to the multitude of closely-allied forms, the dispa- 

 rity in pattern and coloration exhibited by the sexes, and the instability 

 of colouring and markings in the females. It is probable that no two 

 lepidopterists would even approach agreement in discriminating the 

 known species, and the mass of the genus must remain in a very 

 unsatisfactory state until careful breeding of successive generations 

 from the ova can be systematically applied to its elucidation. Even in 

 South-Africa, as far as I can learn, only two or three species have been 

 reared from the larvae ; but some little aid has been afforded by the 

 record of the capture of the sexes in cojiuld. 



In Mr. Butler's "Revision" in 1876, forty-one South- African 

 species were enumerated, and of these nineteen were for the first time 

 described. I have had the advantage of examining the tjq^es of these 

 species in the very large series of Tvracohis contained in the collection 

 of the British Museum, and comparing them with my own series from 

 South Africa, and arrived at the conclusion that only five of them 

 presented characters warranting specific separation from previously 

 desci'ibed forms. 



The beauty of most of these butterflies is very remarkable, espe- 

 cially in the $ s, wdiere the brilliantly-tinted tips of the fore-wings, 

 usually relieved by black edging, contrasts with a pure-white or pale- 

 yellow field. The lone group perhaps carries off the palm of loveli- 

 ness, the lustre of the glittering violet tips (in T. Regina quite metallic) 

 being unequalled in any other group, though the intense rich crimson 

 of the tips of T. Danae and its near allies is almost equally splendid. 

 Bright-red, orange-red, orange, and yellow of various shades are the 

 colours that ornament the wing-tips of most of the groups, while the 

 Celimcne section has an immense purple patch of a lustre inferior only 

 to that in the lone section. 



As will be seen, however, in the account of the characters of the 

 several sections given above, there are five groups, viz., those of 

 Manctnliari, Halimedc, Fausta, Amata, and Chrysonome, in which the 



