MOEPHOLOGY OF THE LEPIDOPTERA. 157 



outer margins of the groups resembled very* closely the pupae in class (1), whereas those 

 in the centre were more and more dusky according as the group contained more or fewer 

 members. There was in fact a regular gradation in the amount of black pigment 

 present. 



(3) A large number of caterpillars which had just ceased to feed were placed at the 

 bottom of an empty wooden hox ; the box was covered with muslin, and over the muslin 

 was placed the lid of a linendraper's box covered with purple-black paper. The dark 

 side was turned to the caterpillars ; the lid overlapped the box by six inches at least on 

 every side, and it was pressed down by a weight. The prisoners were thus in absolute 

 darkness. Without exception they suspended themselves to the muslin, and the pupae 

 to which they gave rise had extremely dark bodies and peculiar sooty-looking wings. 

 The golden spots at the base of the abortive spines were scarcely visible *. 



These experiments on colour-variation, which Mere quite by the way, seem to me to 

 bear out Mr. Poulton's conclusions, formulated in his paper on "The cause and extent of 

 a Special Colour-relation between certain exposed Lepidopterous Pupae and the surfaces 

 which immediately surround them," in the ' Philosophical Transactions,' vol. 178, 1888, 

 p. 311 et seqq. 



Turning now to the development of the azygos oviduct with its accessory organs, I 

 shall divide their history into the following five stages: — (1) the full-grown caterpillar; 

 (2) the formation of the azygos oviduct, which occurs during the quiescent period 

 preceding pupation ; (3) the stage during the time in which the bursa copulatrix is still 

 in direct union with the oviduct; (4) the stage in which it acquires its adult indirect 

 union ; (5) the formation of the terminal papilla of the abdomen. This mode of division 

 seems to me better than taking stages of so many days old. As a matter of fact different 



* Hy friend Mr. Mitchell was rearing caterpillars of V. urticce at the same time that I was rearing those of V. Io. 

 He fed them in a dark brown box with a few holes punched in the lid for the sake of ventilation. The holes were 

 small and could have admitted but little stray light. The caterpillars turned to pupae which were uniformly of a 

 very dark hue with a slightly reddish ground-tint. Among the nettles brought to me I found two pupae of V. 

 urticce of the most brilliant golden and green hue, so golden and green, in fact, that my friend Mr. Poulton declared 

 them both to be ichneumoned. He was right as to one, and wrong as to the other, which produced a perfect 

 butterfly. 



The difference noted in the text in the tone of the green colour of the ccelomic fluid, in accordance with which the 

 pupae became light or dark, was very striking. Another point worth recording was the presence, at least here and 

 there, of a red pigment in the hypodcrmis (as I suppose) of the caterpillar preparing to pupate, but not visible, 

 however, in the newly formed pupoe, and the brilliant ruddy hue of the residue of food in the digestive tract. It 

 has struck me that an examination of the following points might yield interesting results in connection with the 

 variability of the colour of pupae of V. Io : — the spectroscopic quality of the light passing through the larval cuticle; 

 the origin, distribution, and fate of the red pigment of the hypodermis ; the effect produced by the evolution of the 

 yellow fat-body which increases so enormously in bulk after the suspension of the caterpillar ; the spectroscopic 

 value, if any, of the yellow colour of the fat-body, and of the green-coloured ccelomic fluid of caterpillars reared in 

 light and darkness. It is probable that an answer to these points, into which I hope to enquire in the course of next 

 summer, may lead to a comprehension of the physical basis underlying the colour-changes of pupae. As all the 

 newly formed pupae of V. Io have much the same tint, and the pigmentation is produced rapidly after the assumption 

 of the pupal state, it seems to me that much must depend on the nature of the nutritive material supplied to the 

 pigment-producing cells by the ccelomic fluid. 



