the colours of certain Lepidoptera. 369 



of the larvae. The estimate of 18 hours must be reduced 

 to about 15, but this latter duration was commonly 

 noticed in the previous observations {I. c, pp. 342, 347, 

 351, &c.). 



At the same time, I should be glad for these conclusions 

 to be tested by the observation of larger numbers, and of 

 many companies. My previous results depended on such 

 varied material, which, upon the whole, gave such dis- 

 tinct testimony in favour of longer stages, that it is 

 possible that some of the difference may be due to the 

 hereditary individual predispositions of the 44 larvae 

 observed in 1887. 



Eelying on the latter observations alone, we should 

 conclude that the preparatory period varies from 20 to 

 24 hours, Stages I. and II. together from 5 to 9 ; while 

 Stage III. has a nearly constant duration of 15 hours. 



Bearing upon the lengths of preparatory stages in 

 (liferent conditions : — 



Under any circumstances the observation has a clear 

 bearing upon the conclusion I had previously arrived at, 

 — that darkness protracts the stages. Whether we con- 

 sider the larvae with or without food, the results are the 

 same : Stages I. and II. are longer in the tin box than 

 in the gilt box, in the dark cylinder than in the tin box, 

 and the pupal colours become darker in the same order 

 (see precedmg table, pp. 366, 367). And the difference is 

 much clearer in the larvae with food, which have been 

 shown above to be, in all probability, in a more normal 

 state. The conclusion previously arrived at was capable 

 of two explanations : darkness might directly protract 

 the stages, or its action might be indirect, tending 

 towards the production of dark pupae, and time being an 

 element in the formation of the superficial pigment, or 

 rather of some colourless precursor. The latter view is 

 strongly supported by the observations here recorded ; 

 for the difference in duration is true of the tin- as com- 

 pared with the gilt-box, in which, although both were in 

 light, there was a corresponding difference in the pupal 

 colours. P'urthermore, the more marked difference, in 

 the case of the larvfe with food, corresponded to a more 

 marked difference of pupal colours, although unattended 

 by any difference of illumination, as compared with the 

 larvae without food. All these statements will be found 



