309 



trical fields. It seems hard to understand the distinction be- 

 tween such bundles and entities generally classed as material. 



In the course of this investigation we have made a num- 

 ber of experiments on the quantities and qualities of the sec- 

 ondary radiations. This subject has been fully treated by 

 Barkla, some of whose recent papers have not yet reached us, 

 and any discussion we gave might be merely a duplication 

 of part of his enquiry. There is, however, one point to which 

 we should like to refer. 



Very hard y rays follow a density law of absorption, 

 treating all atoms alike, except in respect to weight. Soft 

 y rays are not independent of atomic groupings of matter, and 

 are far more strongly absorbed by heavy atoms than by light 

 after allowance has been made for weight. The same is gene- 

 rally true of X-rays ; but in the case of very soft X-rays 

 there is a tendency to revert to the density law again. For 

 instance, X-rays that have passed through the glass of the 

 bulb are soft to copper, silver, tin, and so on, 

 but hard to aluminium, carbon, and low atomic 

 weight generally. No doubt those rays which are 

 soft to such light atoms have already been absorbed 

 by the glass. But secondary X-rays from most sub- 

 stances are softer than anything emerging from the bulb 

 and contained in the primary ray. The difference is not very 

 great when the absorption is measured with the aid of screens 

 made of substances of the higher atomic weights, because to 

 these the primary rays are soft already. But if the screens 

 are made of aluminium, still more of filter paper, the differ- 

 ence now seems to be very great, for the secondary rays are 

 soft even to low atomic weights. For example, in one experi- 

 ment a sheet of copper weighing 018 gr. per square cm. 

 caused a drop of '401 in the logarithm (to base 10) of the 

 primary rays, and only of 447 in the case of the emergence 

 secondary rays from, copper, of '645 in the case of platinum 

 rays, and '805 of iron rays. But when four filter papers 

 weighing *02 gr. per square cm. were used as screen, the drop 

 in the case of the primary rays was '010 — only one-fortieth 

 of the drop caused by a copper screen of nearly equal weight. 

 In the case of the secondary rays, however, the same screen 

 caused a drop in the case of copper rays of '100, platinum 

 rays 053, and iron rays of '188 — that is to say, for these 

 soft rays the filter papers are much more nearly on an equality 

 with copper, weight for weight, than they were for hard rays. 

 It is interesting to bear this in mind when considering the 

 very large quantities of secondary ionization which some sub- 

 stances seem to give. The ionization is always measured in 

 air, which of course consists of atoms not very different in 



