April 13, 1899] 



NA TURE 



571 



of the green-and-white underside of the dimorphic Aiithocharis 

 belia to the respective resting plants of each season is also 

 indicated. 



The poverty, however, of such instances among the seasonally 

 dimorphic species of the European butterfly-fauna is manifest ; 

 and it is thus satisfactory to find Weismann turning, in support 

 of his view, to the numerous striking cases (first brought to his 

 notice in 1894 in a paper by Dr. G. Brandes) of seasonal 

 dimorphism occurring in tropical and subtropical regions, 

 among which were instances where one seasonal form at least 

 assumes a special protective colouring. Hitherto all the cases 

 investigated and experimented on, whether in Europe or North 

 America, had been found referable to the influence of high and 

 low temperatures, and nobody seems to have suspected the 

 occurrence of similar seasonal variation in hot countries ; but, 

 as Mr. L. de Niceville, Mr. W. Doherty, and other observers 

 have pointed out, and as Weismann was apt to recognise, the 

 alternation of wet and dry seasons is as actively inciting an 

 agent in the production of seasonal dimorphism in many parts 

 of the tropics, as that of hot and cold ones is in the temperate 

 latitudes. 



I must confess that I shared in the prevalent erroneous 

 opinion that seasonal dimorphism was not to be looked for in 

 countries without summer and winter seasons of greatly differing 

 temperatures ; and no doubt this was mainly due to my never 

 having resided for any length of time in a region where the 

 rainy season is the warmer and the dry one the cooler. In the 

 south-west of the Cape Colony, where I was stationed, exactly 

 opposite conditions prevail, and in the rainy winter, scarcely a 

 dozen species of butterflies appear, and none of them presents 

 any marked difference from the dry summer specimens of the 

 same species. I was thus unprei:)ared to attach due value to 

 the suggestion, by my friend, Mr. W. D. Gooch, as early as the 

 year 1877, of the occurrence of differing seasonal forms of 

 butterflies in Natal, or to the opinion to the same effect given 

 by Mr. A. J. Spiller in 1880 (Enlomologisl, vol. xiii. p. 3). I 

 believe this communication of Mr. Spiller's to have been the 

 first published information of the apparent occurrence of 

 seasonal dimorphism in the warmer parts of the world ; and 

 the four cases which he specially notices (in the genera 

 Anthocharis [ — Tera(nlus\ Picris, Myialesis, and Hynanis) 

 are undoubtedly true ones. Mr. Gooch {op. eit,, pp. 226 and 

 273) published his concurrence in the main with iMr. Spiller's 

 view, but at the same time mentioned that, in the only two 

 attempts he made to test the matter, by rearing Teracohis 

 onipkalc and Pieris severina^ he found no difference between 

 the winter and summer broods, both belonging to the theoretical 

 winter form with reduced black markings. 



It was in 18S5 that Mr. L. de Niceville, the well-known 

 authority on Indian butterflies, published' a notice of apparent 

 seasonal dimorphism in several species of Calcutta Satyrinae 

 of the genera Mycahsis, Ypthiiim, and Melanitis — the wet- 

 season form presenting distinct ocellated spots on the underside, 

 and the dry-season form being without those markings. He 

 suggested as a possible explanation, that while the conspicuously 

 marked wet-season form is concealed by the dense vegetation, 

 the dry-season non-ocellated form had in the scantily-clothed 

 jungle found protection by the gradual loss through natural 

 selection of the conspicuous markings. Mr. de Niceville's 

 specimens illustrating his paper were exhibited at a meeting 

 of this Society in February 1S85, but his view did not meet 

 with much acceptance among the members present, nor was any 

 alternative explanation of the phenomenon brought forward. 

 He was able, however, in the following year to adduce proof 

 of the correctness of his theory in a memoir - giving details of 

 the rearing of one seasonal form from eggs laid by the other in 

 four of the seven cases named by him in his previous paper, 

 viz. Ypthima hubneri and Y. howra : Y. philomda and }'. 

 iiiarshaltii ; Alycalesis mineiis and M. iiidistaiu ; Melanitis 

 leda and M. ismene ; these pairs consisting respectively of the 

 ocellated wet-season form and non-ocellated dry-season form of 

 each species concerned. 



Just previously to the latter notable record of Mr. de Nice- 

 ville, Mr. W. Doherty had contributed to the same Journal ^ 



1 " List of the Butterflies of Calcutta, &c." (Journ. Asiat. Sue. Bengal, 

 liv. pi. ii. p. 39.) 



-"On the Life-History of certain Calcutta Species of Satyrinae, with 

 special reference to the Seasonal Dimorphism alleged to occur in them." 

 (Of. cit., Iv. pi. ii. p. 22g, iS36.) 



:> " A List of Butterflies taken in Kumaon." IJourn. Asiat. Soc. 

 Bengal, Iv., pt. ii. p. 107.) 



NO. 1537, VOL. 59] 



his four years' observation of seasonal variation while collecting 

 Indian butterflies. He brings to notice that, speaking generally, 

 there were fotir broods annually in that country, viz. two in 

 the wet season and two in the dry season, and that, while there 

 was no perceptible difference between the two broods of the 

 same season, there were often very marked dilTerences between 

 the wet-season broods and the dry-season ones. 'These differ- 

 ences included size (the wet-season form being usually smaller), 

 the angulation of the wings, and the colouring and ocelli 

 of the underside, and were well illustrated by species of 

 Jimonia, Ypthima, Alycalesis, and Melanitis. The author 

 remarks that some countries with wet climate do not yield any 

 but wet-season forms,' and conversely that some very dry 

 countries produce only dry-season ones, instancing the case of 

 yitnonia almana, the dry-season form of which alone occurs in 

 Scinde, while its wet-season form (asterie) only is met with in 

 Ceylon and Singapore. He is of opinion that De Niceville's 

 view is strengthened by the fact that the dry-season forms are 

 more or less leaf-like both in shape and in the underside colour- 

 ing, while no such resemblance is manifested by the wet-season 

 ones, and argues that this points to the greater exposure to 

 danger in the dry season ; but he is inclined to think that the 

 eye-like underside markings in the wet season may serve as a 

 protection from the attacks of birds. It is singular that, while 

 this observant collector enumerates no fewer than twenty-three 

 species of Pierinae in his "List," he does not seem to have 

 noticed the occurrence of seasonal dimorphism in the subfamily 

 which is especially fertile in illustrations of it. 



In view of the satisfactory evidence afforded by De Nice- 

 ville's experiments with Indian Satyrina;, I could no longer 

 doubt that many hitherto puzzling cases of variation among 

 African butterflies would find their solution in the same way, 

 especially as the dated specimens accessible all pointed to the 

 seasonal character of the varieties. I kept the question con- 

 stantly before my entomological correspondents in Natal and 

 the other warmer parts of South Africa, and was enabled by 

 their assistance to indicate in 1889 (" South--'\frican Butter- 

 flies," iii. pp. 6, 7, 125, and 395, 18S9), various extremely 

 probable instances of a corresponding phenomenon among 

 African Satyrinie and Pierinte. Among a most interesting 

 collection made by Mr. A. W. Eriksson in tropical South-west 

 Africa, described by me in 1891 (Proc. Zool. Soc. Lonci., 1891, 

 pp. 59, 64, 85, 89, 96, 97, and 99), I noted what appeared to 

 be undoubted cases of seasonal dimorphism in species of 

 Acrjeinfe, Lycjenidfe and Pierinae ; and again, in cataloguing 

 Mr. F. C. Selous's Manica butterflies in 1894 {op. eit., 1894, 

 pp. 14, 22, 29, 37, 64, and 67), I showed reason for recognising 

 the prevalence of the same kind of variation, especially pointing 

 out how in the case of Melanitis leda all the dated South African 

 examples went to confirm De Niceville's experience at Calcutta, 

 and what strong similar ground existed for considering the much- 

 discussed variation in the Nymphaline Hanianiimida daedaltis to 

 be seasonal. 



An important contribution to the elucidation of the subject 

 was made in 1894 by the late Captain E. Y. Watson in a paper 

 entitled "Notes on the Synonymy of some Species of Indian 

 Pierina"" {Journ. Bombay Nat. Hist. Soc., viii. p. 489 (1894). 



According to this experienced entomologist's observations 

 some species — Terias hecahe, for instance — produce successive 

 broods (from four in the cooler to ten or twelve in the warmer 

 districts) throughout the year, and the last alone of the wet- 

 season or dry-season broods respectively yields offspring exhibit- 

 ing the oppo.site seasonal form ; but it is at the same time 

 pointed out that "in some cases the eggs laid by one female 

 would produce more than one form, according to the state of the 

 atmosphere .shortly before the emergence of each individual, 

 which is the period at which it would be chiefly affected." The 

 author calls attention to the fact that " in different parts of the 

 Indian Region, the seasons vary to a certain extent, so that it 

 cannot be laid down that specimens captured in any particular 

 month will belong to any particular form " ; he defines, how- 

 ever, roughly the limits of the rainy and dry seasons and states 

 that "the very large majority of specimens obtained during 

 those periods will be wet- and dry-season forms respectively." 

 Emphasis is laid on another important point, viz. that the 



1 Mr. de Niceville has recorded (Journ. Asiat. Soc. Bengal, Ixiv., pt. 

 ii. p. 362, 1895) that in N.E. Sumatra r.iin falls in every month of the year, 

 and it is rare for a week to pass without a shower, and that consequently 

 there are no dry-season forms of butterflies to be found there, with the 

 solitary exception of the dry-season form of Melanitis leda, which (as in 

 Java) prevails all the year round as commonly as the wet-season form. 



