112 Professor Forbes's Researches on Heat. 



For heat of 212° the per-centage was still higher, as will af- 

 terwards be shown. 



30. Metallic Gratings. — If the mere defect of transparency 

 were the cause of the peculiar action of scratched surfaces, 

 we might expect that any opake filaments would act in the 

 same way. Could we dispense with the medium altogether, 

 and employ a screen, which should have the qualities which 

 we had artificially given to the physical surface of the medium, 

 we should evidently have advanced a step in the interpretation 

 of the phenomena. The action of grooved surfaces and gratings 

 upon light suggested so forcible an analogy, that before I was 

 able to procure the mechanically striated surfaces, described in 

 thelast article, I had employed fine metallic wire-gauze as a dif- 

 fraction-screen, hoping to obtain results similar to those which 

 I anticipated, and afterwards did obtain, by drawing fine lines 

 upon rock-salt. 



31. The fact that diff'raction-phaenomena in light, produced 

 by gratings, are wholly irrespective of the nature of these gra- 

 tings, as, for instance, whether they be formed of metal-wires, 

 or mere lines drawn through a soapy film stretched on glass, 

 gave some countenance to this experiment. I was not un- 

 aware that diffraction spectra are produced, not by a parallel 

 beam of light, but by a picture, formed of a distant luminous 

 point. Still, though the ground or field illuminated by 

 parallel rays passing through a grating must evidently have a 

 uniform tint, it does not appear absurd to suppose that that 

 tint may be different from white. Nor does this question ap- 

 pear to have occurred to mathematicians or optical writers, 

 until the problem presented itself to me in the course of this 

 investigation. 



32. With such wire-gauze as I could easily procure, I failed 

 in obtaining any peculiarity of action as relates to heat from 

 different sources ; and further, the quantity of heat intercepted 

 by the metallic grating appeared to be nearly, or exactly, pro- 

 portional to the surface of the opake portion of the screen. 

 Thinking that perhaps finer gauze than that I used (60 wires 

 to the inch) might produce the desired effect, I obtained, 

 through the kind assistance of Sir John Robison and M. 

 Leonor Fresnel, the finest manufactured in Paris, going as 

 high as about 160 per inch. In general my first results were 

 confirmed, viz. (1.) that the proportion of heat stopped is ir- 

 respective of the source; (2.) that it is to the incident heat as 

 the area of the wires is to the area of the surface. It must be 

 observed, however, that the determination of this latter pro- 

 portion with extreme accuracy by an examination of the gra- 

 ting, is not so easy as might at first sight appear. When the 



