FOURTH SERIES. CONCLUSION. 25 



ever, has, I believe, first succeeded in integrating the expression for the illumina- 

 tion of a screen placed behind a grating of any kind (See AIHY'S Mathematical 

 Tracts, page 328) on which a plane wave falls, and he informs me, that in every 

 case where the breadth of the interstices is any multiple of the breadth of the 

 wires or opaque spaces, the intensity is the same as if there were a diaphragm 

 equal in size to the sum of the interstices of the grating, 



61. This result (which seems quite sufficiently general for our purpose) is so 

 far confirmed by the absolute indifference of metallic gratings to the quality of 

 the incident heat. 



62. It remains, however, to be explained how furrowed surfaces can act, 

 except by intercepting, as an opaque network would do, a part of the heat. I 

 cannot give an explanation which appears full and satisfactory, but the condition 

 of mica split into thin lamina? by heat, and producing the same effect, may serve 

 to guide us, perhaps, to something like the true cause. 



63. A number of thin plates, of exactly uniform thickness, would transmit 

 a certain colour, and reflect the complementary one. If there be a great prepon- 

 derance of plates approximating to a certain thickness, and if the disproportion 

 of the lengths of the incident waves be great, a large proportion will be in like 

 manner transmitted, and the remainder stifled or reflected. If this effect is not so 

 frequently observed in bodies mechanically separated into films as we might ex- 

 pect, this is owing to the small range of length of wave in the visible parts of the 

 spectrum. A small variation in the thickness of the film transmits or annihilates 

 by interference each colour of the spectrum in succession. If the waves of heat 

 be much more heterogeneous (as I have already surmised) than those of light, 

 such effects would be proportionably more sensible. 



64. Possibly a grooved surface may be considered as presenting a number of 

 polished surfaces, partially detached from the general surface, under small obli- 

 quities to the incident rays ; and we may suppose that these rays, after separa- 

 tion by partial reflection and refraction, reunite with unequal retardations, pro- 

 ducing first a destructive effect upon the shorter waves, and suffering the others 

 to persevere. I have already adverted to the fact, that most turbid fluids trans- 

 mit chiefly the longer luminous waves. I offer these, however, but as vague con- 

 jectures upon a very obscure subject. I think that experiments on the Colour of 

 media, such as those we have employed, and especially of depolished plates, 

 might not be without value in illustrating the phenomena of Absorption in Optics. 



65. In conclusion, it might perhaps be expected that I should take some notice 

 of the experiments and reasonings of which M. MELLONI has addressed an account 

 to M. ARAGO, in two letters dated the 4th and 14th of March last, and published 

 in the Comptes Rendus for the 30th of the same month. These letters were 



VOL. XV. PART I. G 



