468 Mr. i'oulton' B fiirtJier experiments upon 



On the whole it is probable that, when the direct white 

 light is cut off by a screen, and in some cases the mixed 

 reflected light reduced by the same means, the larvae, 

 when resting on reflecting surfaces, are sensitive to a 

 larger part of the spectrum, comprising the red on the 

 one side, and the green on the other, but not the blue. 



In talking the matter over with Sir John Conroy, he 

 suggested that the heating effects of the rays may have 

 something to do with their power; for he informed me that 

 the usual opinion as to the superior heating properties 

 of the red and ultra- red rays is mistaken, and only 

 holds for dispersed light, when the smaller refrangibility 

 of these rays leads to crowding in a given area. Under 

 the conditions of these experiments, he tells me that the 

 yellow rays possess the greatest heating power. 



But if it were a question of temperature, it is very 

 difficult to understand why the effects of reflected light 

 are not completely subdued by those of direct light. Nor 

 is there any evidence that accessory conditions which 

 must greatly affect the temperature of the larvae, such as 

 the amount of sunshine, have any influence upon the 

 result. Upon the whole it appears far more probable 

 that nerve-terminations in the skin are directly affected 

 by the radiant energy, and, in most cases, are especially 

 sensitive to those vibrations which appear to affect 

 animal life most powerfully. 



Some conclusions from the experiments on larvce. — 

 I have here brought together some of the chief results 

 of the shorter experiments. 



Eegularly dimorphic forms, with intermediate varieties 

 rare or wanting, are never, as far as our present know- 

 ledge extends, susce[)tible to surrounding colours, while 

 variable species tend to be so. In this respect Gcometra 

 papilionaria is very interesting, being susceptible when 

 variable during its youth, but not in advanced stages, 

 when it is dimorphic. Among the Geometra, so many 

 of which are strongly susceptible, we meet with well- 

 marked dimorphism in the genus Ephyra, which is 

 apparently not affected by surroundings. 



Noctiue are far less sensitive than Ge»mctr(e, both in 

 relative numbers and in the effect produced in the most 

 marked cases. The most susceptible Noctace, the Catoca- 

 lidce, are purely arboreal forms, like the majority of 



