the colours of certain Lcpidoptera. 467 



Eesults with P. brassica much the same as with the 

 yellow glass, but more irregular, and on the whole darker. 



Green glass. — " Transmitted light. Green. Light of 

 wave-length greater than 572 and less than 517 strongly 

 absorbed." — Eeceptacle LXXVII. 



PupsG of V. 10 formed upon a reflecting background 

 behind this screen, which is far more complete than the 

 last, were uniformly green. Pupte of F. urtlcoi formed 

 behind it upon a feebly reflecting background were dark. 

 Pupas of P. hrassicce under the former conditions were 

 intermediate. 



Pale blue glass.—" Transmitted light. Some red and 

 3'ellow, the yellow-green, and most of the blue. The 

 extreme red below 645 absorbed, a faint absorption-band 

 from 605 to 584, and a slight continuous absorption above 

 548. "—Receptacles LXXXVI. to LXXXIX. 



When in front of backgrounds which reflected all 

 transmitted light, except the blue, intermediate pupse 

 were produced ; wdien with a background which reflected 

 only the red, the pupae were much darker. 



Blue gelatine. — " Transmitted light. Some red, and 

 the yellow-green and blue. Slight general absorption 

 of the red, and an ill-defined dark band from about 608 to 

 578."— Receptacle LXXIV. 



Intermediate pupse of V. io were formed behind this 

 screen. 



Blue cobalt glass.— " Transmitted Hght. The extreme 

 red; yellow-green; a small amount of green, and the 

 blue. Strong absorption of light of wave-length between 

 702 and 576, and also of wave-length greater than 553 ; 

 the upper limit of this strong absorption too ill-defined 

 to be measured." — Receptacles XCI. to XCIV. 



With certain exceptions, dark pupee of P. brassicce 

 were formed behind this blue screen upon reflecting and 

 non-reflecting backgrounds alike. 



Blue cobalt glass. — "Very similar to the above, but it 

 transmits rather more light." — Receptacle XC. Recep- 

 tacle LXXIII. was (at one time) covered with this glass, 

 or that last described. 



Same efl'ects upon P. brassicce. 



The results, upon the whole, supported the argument 

 ah-eady given. V. io is evidently far more sensitive than 

 P. brassicce, and I greatly regret that so few of the latter 

 were subjected to the conditions described above. 



