118 CORRELATION OF PHYSICAL FORCES. 



any others ; the other colours or shades of colour sink the 

 more deeply in proportion as they absorb or cause to disap 

 pear the more light, until we come to the white cloth, which 

 remains upon the surface. The heating powers of different 

 colours are, however, not by any means in exact proportion 

 to the intensity of their light as affecting the visual organs. 

 Thus red light, when produced by refraction from a prism of 

 glass, produces greater heating effect than yellow light in the 

 phenomena of absorption, as has been observed by Sir W. 

 Herschel. The red rays appear, however, to produce a dy 

 namic effect greater than any of the others ; thus they pene 

 trate water to a greater depth than the other colours ; but, 

 according to Dr. Seebeck, we get a further anomaly, viz. 

 that when light is refracted by a prism of water the yellow 

 rays produce the greater heating effect. The subject, there 

 fore, requires much more experiment before we can ascertain 

 the rationale of the action of the forces of light and heat in 

 this class of phenomena. 



In a former edition of this Essay, I suggested the follow 

 ing experiment on this subject : Let a beam of light be 

 passed through two plates of tourmaline, or similar sub 

 stance, and the temperature of the second plate, or that on 

 which the light last impinges, be examined by a delicate ther- 

 moscope, first when it is in a position to transmit the polar 

 ised beam coming from the first plate, and secondly when it 

 has been turned round through an arc of 90, and the polar 

 ised beam is absorbed. I expected that, if the experiment 

 were carefully performed, the temperature of the second plate 

 would be more raised in the second case than in the first, and 

 that it might afford interesting results when tried with light 

 of different colours. I met with difficulties in procuring a 

 suitable apparatus, and was endeavouring to overcome them 

 when I found that Knoblauch had, to some extent, realised 

 this result. He finds that, when a solar beam, polarised in a 

 certain plane, is transmitted perpendicularly to the axis of a 



