1883.' 



on the Size of Atoms. 



187 



adhere however to one simple system, for it is only because we are 

 in England that the yard and inch are put before you at all, among 

 the metres and centimetres. You see on the diagram then the metre, 

 the centimetre, the millimetre, with circles of the same diameter. 

 Somebody tells me the millimetre is not there. I cannot see it, but it 

 certainly is thei'e, and a circle whose diameter is a millimetre, both 

 accurately painted in black. I say there is a millimetre, and you 

 cannot see it. And now imagine there is 1-lOth of a millimetre, and 

 there 1-lOOth of a millimetre and 1-lOOOth of a millimetre, and there 

 is a round atom of oxygen 1-1,000,000 th of a millimetre in diameter. 

 You see them all. 



Now we must have a practical means of measuring, and optics 

 supply us with it for thousandths of a millimetre. One of our 

 temporary standards of measurement shall be the wave-length of 

 light ; but the wave-length is a very indefinite measurement, because 

 there are wave-lengths for different colours of light, visible and in- 

 visible, in the ratio of 1 to 16. We have, as it were — borrowing an 

 analogy from sound — four octaves of light that we know of. How far 

 the range in reality extends above and below the range hitherto 

 measured, we cannot even guess in the present state of science. The 

 table before you (Table I.) gives you an idea of magnitudes of length, 



Table I, — Data for Visible Light. 



Line 

 of Spectrum. 



A 

 B 

 C 



D. 



E 



6 



F 



G 



H, 



Wave-length 

 in Centimetres. 



604 X lo- 

 se? 



562 

 895 

 889 

 269 

 188 

 861 

 307 

 968 

 933 



5» 



Wave Frequency, 

 or Number of Periods! 

 per Second. I 



395 

 437 

 457 



509 



570 



617 



697 

 756 



X 10'2 

 3 



7 





■63 6 



»' 





and again of small intervals of time. In the column on the left you 

 have the wave-length of light in fractions of a centimetre ; the unit 

 in which these numbers to the left is measured is the 1-1 00,000th (or 

 10"^) of a centimetre. We have then, of visible light, wave-lengths 

 from 7J to 4 nearly, or 3*9. You may say then roundly, that for the 

 wave-lengths of visible light, which alone is what is represented on 

 that table, we have wave-lengths of from 4 to 8 on our scale of 

 l-100,000th of a centimetre. The 8 is invisible radiation a little 

 below the red end of the spectrum. The lowest, marked by Fraun- 



