March 4, 1909] 



■NATURe 



is partlv addressed, and render it unsuitable for con- 

 tinuous reading except to those already well versed 

 in organic chemistry. But to those who desire to 

 find the latest authoritative information of a chemical 

 kind it will prove an excellent work of reference. 



W. D. H. 



Ccomctrv. Theoretical and Practical. Part iii. By 

 W. P. Workman and A. G. Cracknell. Pp. ix+66. 

 (London: \V. B. Clive, 1908.) Price li. 6d. 

 Tins part of Messrs. Workman and Cracknell's text- 

 book deals with the subject-matter of Euclid, book 

 xi., on modern lines, and contains also an elementary 

 account of the parallelepiped, sphere, and tetrahedron. 

 The characteristics of previous parts are well main- 

 tained ; the brevitv of treatment and the conciseness 

 of arrangement will appeal specially to examination 

 candidates. 



LETTERS TO THE EDITOR. 

 [Ibe Editor does not hold himself responsible for opinions 

 expressed by his correspondents. Neittier can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or any other part 0/ Nature. 

 No notice is talien of anonymous communications.] 



The 7 Rays of Uranium. 



Our knowledge of the y raj's of uranium has until now 

 been confined to their discovery by Rutherford (Pliys. 

 Zcit., 1902, 517) and to the observations of Eve [ibid., 

 1907, iii5). The latter directed attention to their extra- 

 ordinary leebleness and to their relatively low penetrating 

 power. Eve found that uranium gives out only about 

 one-tenth as much 7 radiation as thorium when examined 

 through 0'64 cm. lead, which is most remarkable, con- 

 sidering that it gives about six times more j3 radiation. 

 Whereas the 7 rays of thorium have the same value for 

 the absorption coefficient as those of radium [A.(cm.)-' = 

 from 0-57 to o>46 over a range of from 064 cm. to 30 cm. 

 of leadj, the uranium 7 rays are far more easily absorbed. 

 Eve gave the value 1-4 for \ for thicknesses of lead 

 between 0-28 cm. and 0^92 cm. He stated that the radia- 

 tion was homogeneous, that the absorption was exponential 

 over this range, and that the rays were practically com- 

 pletely absorbed in i cm. of lead. He worked with 

 uranyl nitrate. 



Having at our disposal 50 kilograms of pure uranyl 

 nitrate, provided by the generosity of a friend in connec- 

 tion with the work of one of us on the parent of radium 

 (N.vruRE, January 28, 366), we have been able greatly to 

 extend and in part to correct the work on the uranium 

 7 rays. By a long sequence of chemical operations, known 

 and new, but based largely on the magnificent chemical 

 work of Sir William Crookes, who discovered the sub- 

 stance (Proc. Roy. Soc, 1900, Ixvi., 409), we separated 

 by far the greater part of the uranium X from the 

 uranium, and obtained it, in the last separation, in the 

 form of films weighing only a few milligrams. The 

 operations absorb about twelve days. Uranium X con- 

 tributes, as first shown by one of us (Trans. Chem. Soc, 

 1902, 860), none of the a rays, but all the 3 rays of the 

 uranium, and, as is to be expected, and as the present 

 work shows, the 7 rays also. These have been found to 

 decay at the same rate as the $ radiation, namely, to one- 

 half every twenty-two days. The initial /3 radiation of 

 the bare preparation lit up an X-ray screen to about the 

 same extent as 7 mg. of pure radium bromide contained 

 in a sealed thin glass tube. The luminosity could be 

 plainly seen in a fully lighted room when the screen was 

 held in the shadow of the observer; but as Eve found, the 

 7 radiation is extraordinarily feeble. It was accurately 

 compared with that from a known quantity of pure radium 

 bromide after passage through 2-5 cm. of lead by means 

 of an electroscope. Under these conditions the uranium X 

 was equivalent to 0-053 mg. of radium bromide. As shown 

 later, it can be calculated that the lead screen cut down 

 the y rays of the uranium X to 20-6 per cent., and of the 



NO. 2053, VOL. 80] 



radium to 55 per cent., of their initial values. From these 

 data, allowing for the decay during the processes of separa- 

 tion, it may be provisionally estimated that the radiation 

 of the two elements, uranium and radium, is about as 

 one to five hundred million when, as in the present case, 

 absorption is eliminated and only the hard 7 rays dealt 

 with. 



Before the activity of the preparation decayed too far 

 we were able to determine accurately the absorption 

 coefficient of the 7 rays in fourteen substances. As Wigger 

 found 'for the 7 rays of radium [Jahrb. Radioakt., 1905, 

 430). the absorption follows a strict exponential law after 

 a certain initial thickness of substance has been penetrated, 

 and the absorption coefficient is very nearly proportional 

 to the density of the substance. Thus for lead between 

 the thicknesses of i cm. and 5 cm. — and for all other 

 substances over corresponding thicknesses — the absorption 

 is within the very small limit of experimental error abso- 

 lutely exponential. The value of the absorption coefficient, 

 A (cm.)-', for lead is 062. In general, for all substances 

 the value of A./d, where d is the density, is about 0055, as 

 compared with o-o2i for the radium 7 rays for thicknesses 

 greater than 2-8 cm. of lead (Wigger). Thus the uranium 

 7 rays are about two and a half times more strongly 

 absorbed than those of radium. 



The conditions of the experiment are of fundamental 

 importance, as thev affect very much the value obtained 

 for the absorption coeflicient. In our experiments the dis- 

 position was in the main similar to that of Wigger, in 

 that the absorbing plates were clamped up tightly to form 

 the base of the electroscope, and the preparation was 

 placed in a definite position beneath. For insulators the 

 upper surface was covered with a thin leaf of aluminium. 

 Whenever practicable, the absorbing plates were all of the 

 same material. Only for light substances, and for one 

 experiment with mercury, was the base of the electroscope 

 a plate of lead as in Wigger 's experiments. Its thick- 

 ness was 1-2 cm. 



Our value for the absorption coetTicicnt is entirely 

 different from that given by Eve, and, indeed, it is a little 

 doubtful what ravs Eve observed. Over the range of 

 thickness of lead he used, from 028 cm. to 092 cm., we 

 find that the rays are not homogeneous, and the ex- 

 ponential law does not hold at all. There is present in 

 relatively great intensity a very much less penetrating 

 radiation, completely absorbed by i cm. of lead, with a 

 value for \ from eight to ten times greater than for the 

 penetrating type. The absorption and magnetic deviability 

 of these rays are under examination. They would have 

 been far less prominent relatively in Eve's measurements 

 with uranyl nitrate than in ours with uranium X, owing 

 to the strong absorption in the former case. It may be 

 mentioned that the existence is to be anticipated of a very 

 soft 7 radiation corresponding to the extremely soft 3 

 radiation of uranium X (Schlundt and IVIoore, Levin, 

 H. W. Schmidt). There appears to be no radiation corre- 

 sponding with Eve's value of \, but then his value for 

 the 7 rays of radium, 0-46, is about as different from 

 Wigger's, 0-24, as his value for the uranium 7 rays, 1-4, 

 is from ours, 062. 



The value found in our experiments for K/d, 0055, was 

 actually obtained exactly for substances so different in 

 density and nature as mercury, lead, aluminium, slate, 

 and jjine-wood, showing the remarkable range of the 

 " density law " in this case. .\t the same time, we do 

 not think it holds strictly, for brass (density 8-40) actually 

 absorbed more than copper (880), and zinc (7-07) more 

 than tin (7-25), in experiments which were strictly corn- 

 parable and under good conditions. The actual experi- 

 mental values of \ld obtained varied within the extremes 

 of 004s (one value for iron) and 0068 (paraflSn wax). 

 Part of this variation, but not, we think, the whole, is 

 doubtless due to experimental error. Although the ex- 

 ponential law holds, so far as we can see, quite strictly, 

 the values obtained for \ appear to depend somewhat on 

 the particular experiment in an as yet not completely ex- 

 plained way. We propose carrying out similar experi- 

 ments with the 7 rays of radium, in the hope of obtaining 

 further light on the nature of the variation. 



Beyond 5 cm. of lead, and corresponding thicknesses of 

 other metals, A appears to change and to become very 



