XXII 



conducted concerning the origin of the pigmental colours in the pupal stage, 

 have established a sequence which does not accord with the one assumed by 

 me and that in these red certainly does not occur as the original pigment. How 

 does the matter really stand in this respect? When the haemolymph of the 

 pupa, penetrated into the scales, has been transformed into a granular pigment 

 it is of a uniform dull yellow or light drab colour. This can, in a certam 

 sense, certainly be termed the oldest pigmental colour but only when this 

 pigment, as Mayer ^) expresses it, appears as "ornamentation", does it acquire, 

 through chemical changes in the haemolymph itself, what he terms the mature 

 colours which, therefore, arise from the action of chemical reagents ; experiments 

 have shown that in this manner, according to the reagents employed, red-orange 

 or yellow can be produced. By the term "mature colours", therefore, is 

 meant the colours which are visible in the imago after it has left the pupal 

 envelop and it is only with these the theory of colour evolution is concerned ; 

 when, in discussing these I used the term original pigment, I referred to the 

 pigment which, in the original forms of Rhopalocera, besides some black pigment, 

 constituted the universal mature colour and which in the present forms has since 

 been more or less modified by the process of colour evolution, but not what 

 has been termed original pigment in the ontogenetic investigations which as 

 such never appears as a mature colour and can only be classed with it after 

 undergoing chemical change. In consequence of what I have just stated con- 

 cerning these mature colours it is, therefore, (juite possible that in the primary 

 butterflies this colour should have been red to become gradually changed in 

 later times into orange, yellow, and white, especially since the pigments of all 

 these colours, in so far at least as this has been investigated, as in the Pieridae, 

 do not differ chemically. These changes must, however, have proceeded very 

 unequally in the different groups of scales which alone can explain the fact that 

 several of these pigmental colours appear side by side on the wings of the same 

 animal. This inequality of its expression in time and space is, moreover, as 

 has been correctly observed years ago by Weissmann, a fixed character in all 

 evolutionary processes and does not prevent these colour changes in any way, 

 from being governed by one and the same principle or, as it is usually termed, 

 by a fixed law. I always hesitate, however, to employ this so usual expression 

 " law " in this connection, since I am only too well aware that non-lawyers 



') I follow here principally the two treatises by Alfred Goi.DnoROUGH Mayer, " The devel- 

 opment of the iving scales and their pigments in butterflies and moths" and" On the color and color- 

 patterns of Moths and Butterflies" {Bulletin of the Museum of Comparative Zoology at Harvard 

 College, xxix, n". 5, xxx, n". 4). Mayer's term " ornamentation " is good in itself, providing it 

 is not understood to imply that the process in question occurs for the sake of ornamentation. 



