ISO 



NATURE 



[June 13, 1907 



LETTERS TO THE EDITOR. 



,lThe Editor does not }wld himself responsible for opinions 

 expressed by his correspondents. Neither can he undertake 

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

 manuscripts intended for this or any other part of Nature 

 No notice is taken of anonymous communications.] 



The Oiigin of Radium. 



I CANNOT let Prof. Rutherford's letter in NatL're of 

 June 6 pass without directing attention to one striking 

 .consequence, in which I personally am interested. During 

 1904 and 1005 I published (Natuke, May 12, 1904, January 

 26, 1905, and P/ii7. Mag., June, 1905, p. 768) the result of 

 an experiment which went to show that a kilogram of 

 uranyl nitrate, purified initially from radium by precipi- 

 tating barium as sulphate in its solution, and kept 550 

 days, generated a quantity of radium which, although only 

 one-thousandth part of what is Iheoretically to be expected 

 on the view that a direct change of uranium X into radium 

 takes place, was still one hundred times the amount 

 initially present. Boltwood {.■Im. Journ. Sci., September, 

 1905, XX., 239), woiking with one hundred grams of uranyl 

 nitrate purified from radium initially by repeated crystal- 

 lisations from water, was unable to observe any detectable 

 increase after a period of ,igo days, and concluded that 

 ■" the results obtained by Mr. Soddy arc without signifi- 

 cance," and averred that my results were due to the in- 

 troduction of radium salts during the tests. 



Now such a criticism and such an imputation on the 

 part of one investigator dealing with the work o'f another 

 surely ought only to have been made if it was the only 

 possible explanation of the discrepancy. As it was, to me at 

 least, it was not even the most obvious explanation. Bolt- 

 wood did not give consideration to the all-important in- 

 fluence of the method of purification of the uranium from 

 radium on the results obtained. My result, that the rate of 

 production of radium from uranium was only one-thousandth 

 of the theoretical, brought into being the present theory of 

 the existence of several hypothetical intermediate transition 

 forms between uranium and radium. It is obvious that, 

 according as the method of purification employed does not 

 or does remove these transition forms as well as the 

 radium, so one will or will not expect to observe an initial 

 production of radium in a solution of uranium. Now the 

 method of precipitating barium as sulphate in a uranium 

 solution is designed to remove only the radium, whereas 

 the inethod of repeated crystallisation from water adopted 

 by Boltwood is well calculated to purify the uranium, that 

 is, to free it from all other accompanying substances. 

 Hence there is no necessary discrepancy between the results 

 of the two experiments. This view has been put forward 

 by Rutherford (" Radio-active Transformations," p. 75q). 



I would not now have raised this matter had not 

 history apparentlv repeated itself, and Prof. Rutherford's 

 most recent results (Nature, June 6, p. 126) enabled me, 

 without making any special claim to infalllbilitv, to exhibit 

 clearly the real nature of Boltwood's criticism. In the 

 American Joutna] of Science for December, IQ06. p. ■^n.-j 

 fNATURE, January -^l, Boltwood published a " Note on the 

 production of Radium from .Actinium " in which evidence 

 was given that actinium is the parent of radium. This 

 was quickly followed (Nature, Januarv 17, p. 270) by some 

 confirmatory evidence of a similar character bv Rutherford, 

 who.^ however, pointed out that there was no proof th,-it 

 actinium was itself the true parent of radium, although 

 this parent was undoubtedlv present in the actinium solu- 

 tions employed. Now Rutherford shows in last week's 

 issue that actinium purified from radium in a different 

 manner yields no appreciable growth of radium. Is Bolt- 

 wood's previous positive result then " without signifi- 

 cance"? Surely not. But if Boltwood's result on the 

 production of radium from actinium can be explained, as. 

 of course, it can be explained, without charging him with 

 introducing radium into h's solntio". so in the same wav 

 can mine with uranium. Indeed, whereas the intermediate 

 product, which Is the parent of radium, is a necessarv 

 companion of anv uranium preoaration which has not 

 recentlv been subjected to a purification process capable of 

 removing It. it has yet to be shown that the association 

 NO. 1963. VOL 76] 



of this parent with actinium is genetic and not purelv 

 fortuitous. 



I hope this exposure of an old criticism, made without 

 due consideration of the complexity of the problem, will 

 clear the way for the publication of some further results. 

 In the two years that have elapsed since the publication 

 of my last paper I have had the advantage of the co- 

 operation of Mr. T. D. Mackenzie in the steady and con- 

 tinuous prosecution of the work under the most favourable 

 conditions. We have from the commencement, which dates 

 prior to Boltwood's first communication on the subject, had 

 as the basis of the work the all-important influence of the 

 method of purification adopted, and we have used through- 

 out a new method of purification, which, though not with- 

 out difiiculty and danger in its application to the purifi- 

 cation of large quantities of uranium salts, was deliberately 

 chosen as affording a reasonable guarantee that it would 

 separate the uranium from all other substances present. 

 Mr. Mackenzie has purified with the utmost care three 

 separate kilograms of uranyl nitrate by this method, and 

 I may anticipate our results to the extent of saying that, 

 so far, they entirely confirm and extend the results obtained 

 by Boltwood in which re-crystallisation was the method 

 of purification employed. The first preparation, contain- 

 ing after purification about 500 grams of uranyl nitrate, 

 has been kept for 600 days, and has not shown the slightest 

 detectable increase in the amount of radium initially pre- 

 sent. Now that these three purified preparations have 

 been set up in a form to allow of continuous and extended 

 observation, our attention is being directed to the residues 

 from the three kilograms, which should contain the parent 

 of radium, if my earlier positive result was correct. After 

 all, it would be a little surprising if this parent of radium 

 was entirely absent from commercial salts of uranium, for 

 although Boltwood and Rutherford have found it in pre- 

 parations of actinium, it must not be forgotten that the 

 only source of actinium is that from which commercial 

 uranium salts are prepared. 



Frederick Soddy. 



The L'niversity, Glasgow, June 8. 



The Structure of the .ffither. 



In the current number of the Philosophical Magazine 

 I have given in some detail certain objections to identify- 

 ing the magnetic vector with translational a;thercal motion, 

 and to a large extent these are on all fours with Prof. 

 Hicks's objection, which is cited by Sir Oliver Lodge in 

 the same number, and of which 1 had lost sight. Very 

 briefly, thus: if bodily a;ther flow were (within a constant 

 factor) identical with magnetic induction, or were even an 

 essential feature thereof, our judgment as to whether or 

 not a given region was pervaded by magnetic induction 

 would depend on the arbitrary origin of coordinate axes 

 relatively to which we chose to measure velocities, motion 

 of bodies through the a;ther being physically indistinguish- 

 able from an equal and opposite motion of the sther with 

 those bodies at rest. 



Much the same diflTiculty (concerning the essential 

 relativity of motion) seems to me to arise when resultant 

 ;Ethereal momentum is taken to correspond to the vector 

 product of the electric and magnetic vectors ; in this case, 

 moreover, further dilliculties are encountered. Consider- 

 ation of a progressive train of electromagnetic waves shows 

 that, with this ;ethereal-flow interpretation of the Poynting 

 vector, we should have a resultant rethereal motion made 

 up of a steady flow in the direction of wave propagation, 

 together with to-and-fro motions parallel to that direction 

 and kinematically exactly simulating the motion of a gas 

 which is transmitting waves of sound. This clearly implies 

 compressibility of the aether, not merely as a minute 

 residual phenomenon, but as a fundamental relation of 

 electromagnetism. 



And what would happen in the case of such a body as 

 the sun, which consistently radiates more energy than it 

 receives by radiation? There w^ould be a flow of aether J 

 outw.ard in all directions, maintained throughout immens 

 periods of time. This difiiculty seems almost insuperable. 



There appears to me to be much evidence in favour ofl 

 the view thai the resultant velocity of the a>ther (referred! 



