372 



side-veins, springing perpendicularly from both sides of the main 

 tracheae, only penetrate for a short distance into the internervural 

 spaces, and end there with a knob formed by a coiled part of 

 the air-tube, the middle part of the space thus remaining free 

 from tracheae, and thereby reflecting the light in a somewhat 

 different way. 



Nevertheless I have satisfied myself, by careful inspection of 

 fresh preparations, that, especially in P. podaHrius, a real 

 difference of colour-shades is perceptible, corresponding with 

 the pattern of the pupal wing-sheath. 



In Pieris hrassicœ- the difference between the central part 

 and the margins of the internervural wing-spaces likewise becomes 

 manifest, but only to a very slight degree, and during a short 

 period preceding the universal whitening of the whole wing. 



The above-mentioned observations on the development of 

 a primitive colour-pattern seem to me to offer a weighty argument 

 in favour of the hypothesis which I founded on the ornamenta- 

 tion of the pupal wing-cases and tried to strengthen by a com- 

 parison of the different imaginai patterns. The hypothesis in 

 question was formulated by mein my article of 1911, " Ueber 

 die Phylogenie der Flügelzeichnung bei Tagschmetterhngen," 

 in this wise : that the primitive wing-pattern of all Rhopalocera 

 was brought about by a difference in coloration between the 

 nervures and the median stripes of the internervural spaces, 

 and that all the complicated imaginai patterns, with their almost 

 infinite diversity of colours and markings, might be derived 

 from this fundamental chstribution of coloured matter over 

 the wing surface. 



Imaginal Forms. 



The examination of the pupal wing-sheaths, and the investi- 

 gation of the young imaginal wings developing within, have 

 led us to one and the same conclusion, viz. the occurrence during 

 metamorphosis of a primitive wing-pattern, common to difierent 

 families of Rhopalocera, and contributing more or less to the 

 elements of the definite or imaginal pattern. 



This primitive pattern we may, in a general way, describe 

 as consisting of pigment-accumulation along the nervures and 

 in the spaces of the wing- area between them, the former as well 



