728 Professor Sir J. J. Thomson [April 14, 



and the other filled with the different specimens of foil ; this plate is 

 raised or lowered by the windlass W, which winds up a piece of silk 

 fastened to the plate. In some experiments the diaphragm itself 

 was used as a target, the cathode being put opposite the diaphragm. 

 This foil of various kinds was fixed in front of the diaphragm ; and in 

 this way the impact of cathode rays against different materials could 

 be investigated. 



In this method the energy of the rays is measured by the inoniz- 

 ation they produce in the chamber. To absorb all the radiation it is 

 necessary to have the gas in this chamber at a higher pressure than is 

 admissible in the tube where the cathode rays are produced ; the 

 chambers must therefore be separated by a diaphragm which can 

 stand a difference in pressure of some millimetres of mercury and 

 yet be thin enough to allow the rays to pass through them. As these 

 rays are extremely easily absorbed this necessitates the use of dia- 

 phragms which are less than a thousandth of a millimetre in 

 thickness. Considerable difficulty was found in getting diaphragms 

 to satisfy these conditions ; the method finally adopted is due to my 

 assistant, Mr. Everett, and consists in putting a drop of a solution of 

 celluloid in amyl-acetate on the top of mercury in a shallow vessel. 

 As the solution evaporates a thin film of celluloid is found on the 

 mercury, and this is allowed to get thoroughly dry before being 

 removed. By this device the boundary is free while the film is 

 drying ; if it is fixed the film tears from the shrinkage which takes 

 place while it is drying. 



I will now consider some of the results obtained by this method, 

 and will first take the change in the absorption of the rays by a layer 

 of aluminium, gold-leaf, or silver as the energy of the cathode rays 

 producing the Rontgen radiation is gradually increased. We know 

 that, speaking generally, hard rays correspond to large energy, soft 

 rays to small, and it might be thought that as we increased the 

 energy we should get a fairly constant rate of increase in the propor- 

 tion of the radiation gettino: throusrh the foil. When we trv the 

 experiment we find, however, that the increase in the transmissibility 

 of the Rontgen rays with the energy of the parent cathode rays is not 

 of this uniform character. The graph representing the relation 

 between the absorption of the Eontgen rays and the energy of the 

 cathode rays much more nearly resembles a series of steps with flat 

 places between them than a unifoi-m slope. Great changes in the 

 energy of the cathode rays may give rise to very small ones in the 

 absorption of the Rontgen rays they produce, as the absorption 

 fixes the character of the Rontgen rays ; this shows that this has 

 altered but little in spite of the large alteration in the energy of the 

 cathode rays. This is illustrated by the following Table, which gives 

 the absorption produced by the same pieces of foil when the voltage 

 of the cathode rays is steadily increased ; the numbers given represent 

 the ratio of the energy in the rays before and after they pass through 



