204 Sir William Thomson [Feb. 2, 



their successive order, elastically connected with it — we shall have a 

 rude mechanical explanation, realisable by the notably easy addition 

 of the proper appliances to the dynamical models before you, to ac- 

 count for refractive dispersion in an infinitely fine-grained structure. 

 It is not seventeen hours since I saw the possibility of this explana- 

 tion. I think I now see it perfectly, but you will excuse my not 

 going into the theory more fully under the circumstances.* The 

 difficulty of Cauchy's theory has weighed heavily upon me when 

 thinking of bringing this subject before you. I could not bring it 

 before you and say there are only four particles in the wave-length, 

 and I could not bring it before you without saying there is some other 

 explanation. I believe another explanation is distinctly to be had in 

 the manner I have slightly indicated. 



Now look at those beautiful distributions of colour on the screen 

 before you. They are diffraction spectrums from a piece of glass 

 ruled with 2000 lines to the inch. And again look, and you see one 

 diffraction spectrum by reflection from one of Rutherford's gratings, 

 in which there are 17,000 lines to the inch on polished speculum- 

 metal. The explanation by " interference " is substantially the same 

 as that which the undulatory theory gives for Newton's rings of light 

 reflected from the two surfaces, which you have already seen. Where 

 light-waves from the apertures between the successive bars of the 

 grating reach the screen in the same phase, they produce light; 

 there, again, where they are in opj)osite phases, they produce darkness. 



The beautiful colours which are produced depend on the places 

 of conspiring and opposing vibrations on the screen, being different 

 for light-waves of different wave-lengths ; and it is by the measure- 

 ments of the dimensions of a diffraction sj)ectrum such as the first 

 set you saw (or of finer spectrums from coarser gratings) that Fraun- 

 hofer first determined the wave-lengths of the different colours. 



I have now, closely bearing on the question of the size of atoms, 

 thanks to Dr. Tyndall, a most beautiful and interesting experiment 

 to show you — the artificial " blue sky," produced by a very wonderful 

 effect of light upon matter, which he discovered. We have here an 

 empty glass tube — it is " optically void." A beam of electric light 

 passes through it now, and you see nothing. Now the light is stopped, 

 and we admit vapour of carbon disulphide into the tube. There is 

 now introduced some of this vapour to about 3 inches pressure, and 

 there is also introduced, to the amount of 15 inches pressure, air 

 impregnated with a little nitric acid, making in all rather less than 

 the atmospheric pressure. What is to be illustrated here is the 

 presence of molecules of substances produced by the decomposition 

 of carbon disulphide by the light. At present you see nothing in the 

 tube ; it still continues to be, as before the admission of the vapours, 



* Farther examination has seemed to me to confirm this first impression ; and 

 in a paper on the Dynamical Theory of Dispersion, read before the Koyal Society 

 of Edinburgh, on the 5th of Marcli, I have given a mathematical investigation of 

 the subject.— W. T., March 16, 1883. 



