LI^•yEA^' SOCIKTY Ol' LONDOX. 37 



b. Tlie Edijes of Marliiujs la tlie I'atterns of Modi/s and tlwir 

 jMhnics. — Witli i-ertaiii exceptions — uuiinportant from our prejsent 

 slaiulpoint — the pigments* of butterflies' wings are contained in 

 the scales. Two colours do not occupy the same scale, but the 

 pasisage from one colour to another is effected by a set of scales 

 containing one pigment being replaced by a set containing another. 

 In the examples referred to by Professor Punnett in the passage 

 quoted on pp. 35, ^6, the yellow scales are suddenly replaced by 

 black scales at the margin of the hind- wing patch of Ainawis 

 ecJieria, while they are not suddenly replaced in the mimic, the 

 mima form of IJypolimnas dubius. A part of tbe area that is 

 wholly yellow in the model contains abundant scattered biack 

 scales in the mimic and so with part of the black area which in 

 the mimic contains plenty of yellow scales. Scales that are 

 yellow in the one are black in the other and vice versa. The 

 difference is real. 



The metaphor is also misleading in another way : it seeks to 

 explain the diiJerence as an incidental outcome of specific con- 

 stitution unconnected with mimicry. But the soft outline is a 

 common characteristic of the miujetic pattern just as the hard 

 outline is of the model's. The Hiipolimnas mimic is far more 

 closely allied to the Dauaine model Amauris ecJieria than either of 

 them are to the Papilionine mimic. P. dardanus, yet both mimics 

 exhibit the soft-edged pattern. They exhibit it in fore wing as 

 well as hind : they exhibit it in all their other forms mimetic of 

 other comparatively hard-outlined Danaine models. The following 

 sentence, published by the writer fourteen years ago, refers to the 

 most conspicuous and best-known black and v\hite model and 

 mimics in Africa : — " . . . The white centre of the upper surface 

 of the wings deepens gradually at its margin into black in both 

 Papilio an(l IVymphalid, while the n)argin of the corresponding 

 white area in the Danaine exhibits an extremely sharp and abru])t 

 trnnsition into black " t. 



These differences between the edges or borders of markings on 

 butterflies' wings are important and 1 pro])Ose for theni the terms 

 eidegnic (with a clear edge), and di/slee/inc (with a dim or bad 

 edge), tbe latter passing into alecjnic (without an edge), kindly 

 suggested to me by Professor Gilbert Murray J. 



c. Modes of Occurrence of Dtjslegnic Patterns in Buttcrfties, — 

 Dyslegnic patterns are not only common in mimics as compared 

 with models, but are also common in females as compared with 



* The same arginiient apph'es to the physical causes of colour here omitted 

 from consideration for the sake of simplicit}'. 



t Trans. Ent. Soc. Lond., 1902, p. 486, n. The passage refers to the liind 

 wings of the western Danaine Amauris niavius and of its two mimics tlie 

 hippocoon female form of Fapilio dardaints and t)0th .sexes of Hji} oliiiioias 

 (E'uralia)anihcdon, also to the corresponding eastern and soutli-eastein lornis — 

 A. niavius dominicanns, hijjpocoon (or hippocoonidcs as it lias bfcii called in this 

 area), and H. trahUwrgi. 



\ Xeyvvf, an edge or border, with the prefixes ev- good, iva- bud, or a- not. 



