282 KELVIN 



it you see a blackened slip of wood with two white marks on 

 it ten inches asunder. The line on which they are marked 

 is placed perpendicular to the line at which I shall go 

 from it. When I look at this salted spirit-lamp I see a series of 

 spectrums of yellow light. As I am somewhat short-sighted 

 I am making my eye see with this eye-glass and the natural 

 lenses of the eye what a long-sighted person would make 

 out without an eye-glass. On that screen you saw a suc- 

 cession of spectrums. I now look direct at the candle and 

 what do I see? I see a succession of five or six brilliantly 

 coloured spectrums on each side of the candle. But when I 

 look at the salted spirit-lamp, now I see ten spectrums on 

 one side and ten on the other, each of which is a monochro- 

 matic band of light. 



I will measure the wave-length of the light thus. I walk 

 away to a considerable distance and look at the spirit-lamp 

 and marks. I see a set of spectrums. The first white line 

 is exactly behind the flame. I want the first spectrum to 

 the right of that white line to fall exactly on the other white 

 line, which is ten inches from the first. As I walk away 

 from it I see it is now very near it ; it is now on it. Now the 

 distance from my eye is to be measured, and the problem 

 is again to reduce feet to inches. The distance from the 

 spectrum of the flame to my eye is thirty-four feet nine 

 inches. Mr. President, how many inches is that? 417 

 inches, in round numbers 420 inches. Then we have the 

 proportion, as 420 is to 10 so is the length from bar to bar 

 of the grating to the wave-length of sodium light. That 

 is to say as forty-two is to one. The distance from bar to 

 bar is the four hundredth of a centimetre: therefore the 

 42nd part of the four hundredth of a centimetre is the wave- 

 length according to our simple, and easy and hasty ex- 

 periment. The true wave-length of sodium light, accord- 

 ing to the most accurate measurement, is about a I7,oooth 

 of a centimetre, which differs by scarcely more than one 

 per cent, from our result! 



The only apparatus you see is this little grating a piece 

 of glass having a space four-tenths of an inch wide ruled 

 with 400 fine lines. Any of you who will take the trouble 

 to buy one may measure the wave-length of a candle 



