380 Rev. Mr. PovielVs Historical Sketch [May, 



was adopted by Dr, Fyfe in his experiments on the illuminating 

 powers of oil and coal gas (Edin. Phil. Journ, No. xxii. p. 367) ; 

 but he estimated the intensity by comparison of shadows ; the 

 objections to which plan have been already adverted to. 



Dr. Brewster has proposed a photometer on the principle of 

 viewing a luminous point through a telescope, and drawing in 

 the eye-tube till its image is expanded into a luminous circle. 

 The intensity of light in this circle diminishes as the square of 

 the distance from the point of distinct vision. (Treatise on Phil. 

 Instruments, p. 47, 206), 1813. 



The method of comparing different lights suggested by this 

 author is that of thus attenuating the light till the disk becomes 

 invisible, in each case ; and then comparing the squares of the 

 distances for the ratio. 



In trying this method, the great difficulty I found to be that 

 of being able to say precisely when the object did actually disap- 

 pear ; thus no reliance could be placed on this method for any 

 thing like accuracy. 



But Dr. Brewster's principle might, with much greater pros- 

 pect of success, be applied to a siinuUatieous comparison of two 

 lights. Two similar telescopes might have their eye-tubes fixed 

 in such a position that the extremities next the eye should be 

 close together, and the object ends diverging to the lights, but 

 only forming a small angle. The lights should be distant, and 

 the displacement must take place in the object tube, the eye-tube 

 remaining fixed. A piece of ground glass should be placed close 

 to the eye-holes ; and the luminous images might be viewed 

 simultaneously as formed upon this ; and the equalization thus 

 effected. From some rough trials, I am inclined to think that 

 this method might be found very convenient, and generally 

 applicable ; and in its principle it is certainly susceptible of 

 more accuracy than any other of the same kind'. A convenient 

 stand and apparatus for fixing the telescopes, and drawing in 

 the tubes, might be readily devised. 



One other principle I may here mention, which was originally 

 suggested by Bouguer, and which, perhaps, may be applicable 

 to this sort of measurement. This is the fact that the ii-iage of 

 a luminous object, formed by receiving the rays through a minute 

 aperture on a white screen, varies in the intensity of its illumina- 

 tion according to the size of the aperture. By contriving a 

 variable aperture, a method might, perhaps, be thus afforded of 

 equalizing two lights ; but I have not subjected this idea to 

 trial. 



It is confessedly doubtful whether any such methods would be 

 applicable, except in those cases where we have the lights under 

 our command; but in the instance of a comparison of different 

 artificial lights, they may, I conceive, be not undeserving atten- 

 tion ; and the difficulty arising from difference of colour in the 



