IRLF 
 
 151 
 
LIBRARY 
 
 OF THE 
 
 UNIVERSITY OF. CALIFORNIA. 
 
 RECEIVED BY EXCHANGE 
 
 Class 
 
The University of Chicago 
 
 Founded by JOHN D. ROCKKFKM.KK 
 
 ON THE RELATION BETWEEN THE RADIOACTIVITY 
 AND THE COMPOSITION OF THORIUM 
 AND URANIUM MINERALS 
 
 A DISSERTATION 
 
 SUBMITTED TO THE FACULTY OF THE OGDEN GRADU ATI- 
 SCHOOL OF SCIENCE IN CANDIDACY FOR THE DEGREE 
 OF DOCTOR OF PHILOSOPHY 
 
 DEPARTMENT OF CHEMISTRY. 
 
 By WILLIAM HORACE ROSvS. 
 
 HALIFAX, N. S. : 
 
 Me ALPINE PUBLISHING COMPANY, LTD. 
 1907 
 
 ' 
 
 ^L/FOf, 
 
The University of Chicago 
 
 Founded by JOHN D. ROCKEFELLER 
 
 ON THE RELATION BETWEEN THE RADIOACTIVITY 
 AND THE COMPOSITION OF THORIUM 
 AND URANIUM MINERALS 
 
 A DISSERTATION 
 
 SUBMITTED TO THE FACULTY OF THE OGDEN GRADUATE 
 
 SCHOOL OF SCIENCE IN CANDIDACY FOR THE DEGREE 
 
 OF DOCTOR OF PHILOSOPHY 
 
 DEPARTMENT OF CHEMISTRY. 
 
 By WILLIAM HORACE ROSS. 
 
 HALIFAX, N. S. : 
 PUBLISHING COMPANY, LTD. 
 1907 
 
ON THE RELATION BETWEEN THE RADIO- 
 ACTIVITY AND THE COMPOSITION OF 
 THORIUM AND URANIUM MINERALS. 
 
 PART I. 
 
 Thorium Compounds. 
 
 It has been shown by McCoy 1 that, when account is taken of the 
 absorption of the a-particles, the specific radioactivity 2 of any pure 
 uranium compound is proportional to the uranium content of the 
 substance. This result was confirmed by Goettsch, 3 who worked in 
 this laboratory on an additional number of pure uranium com- 
 pounds. It has been shown 4 that a proportionality likewise exists 
 in the case of uranium minerals, free from thorium, between the spe- 
 cific activity of the mineral and the percentage of uranium con- 
 tained in it. In the case of the minerals, however, the ratio of 
 specific activity to uranium content was found to be 4.15 times as 
 great as for the pure compounds. The excess of activity of the ores 
 is due to radium and its active products. Since the ratio of radium 
 to uranium in minerals is always found to be a constant, this is con- 
 sidered good evidence that radium is a transformation product of 
 uranium. 
 
 The work herein described was undertaken with a view to mak- 
 ing a corresponding study in the case of thorium minerals. The 
 importance of an investigation of this kind is evident on account of 
 the many conflicting statements in the literature respecting the 
 radioactivity of thorium and its different minerals. Schmidt, 5 
 Strutt, 6 and Mme. Curie, 7 have reported that all the thorium min- 
 erals examined by them show radioactive properties. In contradic- 
 tion to this, it was announced by Hofman and Zerban 8 that an inac- 
 
 1 Ber. d. chem. Ges., 37, 2641, (1904). J. Amer. Chem. Soc., 27, 391, (1905). 
 
 2 The specific activity of a radio-active substance is the total activity of 1 
 gram of the substance expressed in terms of the activity due to 1 sq. cm. of a film 
 of uranium oxide sufficiently thick to be of maximum activity. As this standard 
 may be reproduced without difficulty when desired, and always with constant 
 activity, it was used as being the most rational standard for the comparison of 
 radio-active compounds. 
 
 8 J. Am. Chem. Soc., 28, 1548, (1906). 
 4 McCoy, Phil. Mag., [6], 10, 176, (1906). 
 6 Ann. d. Phys., 65, 141. (1898). 
 
 6 Proc. Roy. Soc Lond., 76, 88 and 312, (1905). 
 
 7 Compt. Rend., 126, HOI, (1898). 
 
 8 Ber. d. chem. Ges., 35, 531, (1902). 
 
tive thorium preparation had been prepared by them from a Brazil- 
 ian Monozite sand, and that thoria separated from certain minerals 
 as, cleveite, samarskite, etc., although active when first prepared, 
 had lost its activity some months after its removal from the ore. 
 Baskerville and Zerban 1 also claimed to have prepared inactive thor- 
 ium from a 'mineral found in South America. 
 
 From their study of thorium compounds Eutherford and Soddy 2 
 concluded that commercial thorium nitrate and the purest thorium 
 nitrate, both give rise to the same amount of thorium emanation. 
 Hahn and Ramsay 3 on the other hand were able to extract from the 
 mineral thorianite a radioactive preparation several thousand times 
 more active than thorium itself, and which gave rise to a proportion- 
 ally larger amount of thorium emanation. This radio active pro- 
 duct called " radiothorium," was shown to precede ThX in this 
 series of radioactive products. The question, however, as to whether 
 or not thorium itself is the first member of the series, or acquires 
 its activity from the presence of a small amount of radiothorium 
 accidentally mixed with it, was still left an open one. 
 
 The methods of determining the specific activity of thorium 
 
 compounds. 
 
 All the thorium minerals which were studied in this research 
 contained uranium, though often in very small quantities. This 
 made it necessary to determine the percentage of uranium in each 
 mineral in order to allow for the uranium activity. The specific 
 activity of uranium in pure compounds was found by McCoy and 
 Goettsch 4 to be 790, while the corresponding value for minerals was 
 found to be 3280, or 4.15 times as great. Knowing the percentage of 
 uranium in a thorium mineral, the activity due to uranium and its 
 products can therefore be calculated. This, subtracted from the 
 total activity of the mineral, gives the activity due to the thorium 
 in the mineral. When this was done, it was found, as was observed 
 for the uranium minerals, that that part of the activity due to thor- 
 ium and its products in the mineral was in every case proportional 
 to the percentage of thorium contained in it. 
 
 !J. Am. Chem. Soc., 26, 1642, (1904). 
 
 2 Proc. Chem. Soc , 18, 120 (I902. 
 
 3 J. de Chim. Phys., 38, 3371 and 3, 617, (1905) 
 
 4 J. Am Chem. Soc., 28, 1556, (1906). 
 
5 
 
 For .the determination of the total radioactivity of a uranium 
 compound, two methods have been employed. 1 Each of the two 
 methods involves the quantitative measurement of the activity of 
 thin films of the substance of known weight, varying from about 1 
 to 20 mg. per sq. cm. Films weighing 20 mg. per sq. cm. show 
 maximum activity. One of the methods 2 consists in finding the 
 
 values of the ratio for thin films of various weights, where w 
 a 
 
 is the weight of the film, and a its observed activity in terms of the 
 unit defined above. By graphic extrapolation, the value of this ratio 
 for an infinitely thin film may be found. This limiting value of 
 
 the ratio is designated as I I . In the case of an infinitely thin 
 
 \a/ 
 
 film,, a = % kiW, where k v is the total activity of unit weight of the 
 thin film. The factor ^ is used since half of the rays are directed 
 downward and are therefore absorbed by the metal plate carrying 
 the film. Hence, the total activity, Jc iy of unit weight of a compound 
 
 is &! = 2/( ) . By definition k^ is the specific activity. 
 
 -"(=) 
 
 Tlie correction to be applied to the observed specific activity of the 
 
 evolved emanation, 
 
 When applying this method to the determination of the total 
 activity of thorium compounds, the method of procedure is more 
 complicated than in the case of uranium compounds, owing to the 
 fact that the former give rise to a gaseous emanation, the activity 
 of which is considerable compared with that of the solid film. This 
 is particularly noticed when measurements are made with the heav- 
 ier films, since the absorption of the emanation by the film itself 
 does not take place to the same extent as occurs in the case of the 
 a-rays. Hence, the observed activity of the thicker films of thorium 
 compounds must be corrected for the effect of the emanation. This 
 was done by the method developed by McCoy and Boss. 3 The 
 proper amount of the compound to be investigated was ground up 
 finely with alcohol in an agate mortar, and then transferred to a 
 small beaker. The beaker was rotated rapidly to bring the particles 
 into suspension and the mixture then poured quickly into a circular 
 
 1 McCoy, J. Am. Chera., Soc., 27, 391, (1905). 
 
 2 McCoy, loc. cit. 
 
 3 Am. J. Sci., 21, 433, (1906). 
 
shallow tin dish, having an area of 39.8 sq. cm. The alcohol was 
 allowed to evaporate under a bell-jar which inclosed also some cal- 
 cium chloride. Each film" prepared in this way, and carefully 
 weighed, was placed in the electroscope case and its activity mea- 
 sured after being allowed to remain in the case five minutes in 
 order to permit the maximum amount of emanation to accumulate. 
 After several observations were made, in order to obtain a true aver- 
 age, the electroscope was again charged, and, as soon as the leaf had 
 reached the beginning of the scale in the subsequent discharge, the 
 film was rapidly withdrawn from the case by means of a thread. 
 The small door of the case opened and closed automatically. At the 
 instant of removing the film, the position of the leaf on the scale 
 was noted, and the stop-watch was started. The leaf continued to 
 move with gradually decreasing velocity for about five minutes, at 
 the end of which period the emanation had almost completely 
 decayed, and the motion of the leaf became very slow and uniform, 
 corresponding to the natural leak of the instrument. 
 
 If D be the number of divisions of the scale traversed by the 
 leaf in five minutes, corrected for the natural leak, and E the initial 
 activity of the evolved emanation, then it was shown 1 that E = 
 .173 D. E is small except for the thicker films. Xow the observed 
 activity, !, of a thorium film is too great on account of the activ- 
 ity, E, of the evolved emanation. But the value a l E, which 
 represents the activity of the solid film alone, is too small because 
 the film has lost a fraction of its emanation and ThB. This loss 
 01 activity is equal to 2 E, since the activity of ThB is approxi- 
 mately equal to that of the emanation when present in equilibrium 
 amount. The total activity of the film is k^v. Therefore the f rac- 
 
 2 E 
 
 tion of the total activity lost bv the film is - - , and the fraction 
 
 A; 1 w 
 
 2 E 
 
 remaining, 1 - -- . But the observed activity of this fraction is 
 K! w 
 
 a\ E. Therefore the true activity, a, of the film, if none of the 
 emanation were evolved, is 
 
 a=l2 E 
 k t w 
 
 1 McCoy and Ross, loc. cit. 
 
The specific activity of thorium dioxide. 
 
 The first compound to be investigated was thorium dioxide. 
 This was prepared from Kahlbaum's purest thorium nitrate 
 by heating at first cautiously, and then strongly in the blast. 
 The purity of the nitrate was established by the analysis 
 of a solution made by dissolving 4 grams in 250 CC of water, 
 25 CC evaporated to dryness in a large platinum crucible, and 
 ignited, gave a residue weighing 0.1750 gram. Portions of 25 CC 
 diluted and then precipitated with ammonia gave upon ignition 
 0.1746 and 0.1748 gram. Finally, portions of 25 CC analysed by 
 NeishV method, gave 0.1751 and 0.1754 gram. The agreement 
 of the weight of the thorium dioxide obtained by the direct evapora- 
 tion of the nitrate solution with the amount obtained after precipi- 
 tation of the thorium shows the oxide to be free from any appreci- 
 able amount of non-volatile impurities. Before any activity mea- 
 surements were made, the oxide was allowed to stand over a month 
 in order that it might attain its maximum activity. 
 
 Table I gives the results obtained with very thin films of sample 
 A of pure thorium dioxide without any correction being applied for 
 effect of the emanation. ID is the weight of the film ; a^ the observed 
 activity of the film in terms of the unit of activity. If the time of 
 discharge 2 o<f the electroscope through a given range of the scale 
 is for the film, and S for the standard film of uranium oxide 
 which had and area of 39.8 sq. cm., then the observed activity of the 
 
 n, . 39-80. 
 
 film is a l = 
 
 u 
 
 TABLE I. Thorium dioxide. 
 
 w 
 
 W di 
 
 1 
 
 2798 41-36 -00676 
 
 2020 30-80 -00599 
 
 1026 20 70 -00495 
 
 0546 12-27 -00444 
 w' 
 
 1 =-00387 
 
 The curve obtained by plotting the values of as ordinates 
 and the corresponding values of w as abscissae is nearly a straight 
 
 1 J Am Chem. Soc., 26. 780, (1904). 
 
 2 Corrected for natural leak. 
 
 2 
 
line. The intercept on the axis of ordinates of the prolongation of 
 
 the curve gave the value of the ratio - for an infinitely thin film. 
 
 j 
 
 (w\ 
 - ) . Since it 
 l/o 
 
 has been shown that the specific activity, Tc l9 of a radioactive com- 
 
 /w\ 
 pound is given by the equation, ki= 2/1 J 9 therefore, the speci- 
 
 _iic, activity of sample A, as determined without taking into account 
 
 2 
 
 '<the effect of the evolved emanation, is equal to =517. 
 
 ' 
 
 That this value approximates very closely the true specific activ- 
 ity of the compound follows from the conclusion drawn by McCoy 
 and Ross, 1 who showed theoretically that in the case of a very thin 
 film of thorium compound, the actual activity of the evolved emana- 
 tion exactly compensates for the deficiency of activity of the solid 
 film, and that therefore the observed activity of such a film is equal 
 to its true activity. Owing to the difficulty of preparing, very thin 
 films of different weights all equally uniform, this value of k^ 
 , instead of being taken as the true specific activity of sample A, was 
 used in determining the correction to be applied to the observed 
 activity, a 1? when measurements were made of the activities of 
 
 ! E 
 
 thicker films. Trnis, the true activity of such films = 1 2 E=a 
 
 517tc 
 
 Table II gives the results obtained with films of sample A of 
 varying weights. 
 
 TABLE II. Thorium dioxide. 
 
 w 
 
 a, 
 
 a 
 
 w 
 a 
 
 E 
 
 w 
 E 
 
 1-8893 
 
 58-80 
 
 56-90 
 
 03322 
 
 2-17 
 
 87 
 
 9182 
 
 55-60 
 
 54-70 
 
 01678 
 
 1-20 
 
 77 
 
 2798 
 
 41-36 
 
 41-00 
 
 00683 
 
 43 
 
 65 
 
 -3020 
 
 33-80 
 
 33-66 
 
 00600 
 
 35 
 
 
 1026 
 
 20-70 
 
 20-66 
 
 00496 
 
 18 
 
 
 -0546 
 
 12-27 
 
 12-26 
 
 00445 
 
 .09 
 
 
 f w \ 
 
 [ - ) = -00387 :-l=-58 
 
 Wo 
 
 1 Loc. cit. 
 
For the lighter films, E was too small to be measured directly. 
 
 From the curve obtained by plotting the values of the ratio -= for 
 
 lii 
 
 the heavier films as ordinates and the corresponding values of w 
 as abscissae, in the same way as described for the 1 values obtained 
 
 in Table I, the value of the ratio - for an infinitely thin film 
 
 mav be found. Then, since E = w/\ ~ I, the value of E for any 
 
 V^/o 
 
 particular film may be determined. E was calculated in this way 
 for the three lightest films in Table II. A new curve plotted from 
 the results given Table II, in the same way as was done for the 
 
 (w\ 
 
 values in Table I, gave for 1^1 the same value as before. There- 
 fore, the true specific activity of sample A is 517. 
 
 The specific activity of a radioactive compound may also be 
 determined in another way. It was shown by McCoy 1 that 
 
 k 2 = - log , where Jc 2 is the absorption co-efficient for 
 
 w 1 x 
 
 unit weight on unit area; s the area of each film, 39.8 sq. cm. ; and 
 x is the ratio of the observed activity of a thin film to that of a 
 film of the same substance of equal area and sufficiently 
 thick to be of maximum activity. In the case of sample A, such a 
 film had an activity of 56.9 (Table II.) Therefore, x f for any 
 particular film of the same compound is equal to the observed activ- 
 ity, &i, divided by 56.9. The co-efficient of absorption, & 2 , differs 
 for different compounds, but is constant for any one compound. 2 
 For the sample A of thorium dioxide, the mean value of /c 2 was 
 
 2 A 1 a 
 
 found to be 178. Then, since Jc l = , therefore Je^ for sample 
 
 s 
 
 A = - - = 509. The mean value of ~ki for the two methods 
 
 f y *o 
 
 is 513. Thorium dioxide contains 87.9 per cent, of thorium. There- 
 fore, the total activity of thorium together with its products in 
 
 513 
 
 sample A 187^-5= 584. 
 
 1 .T. Am. Chem. Soc., 27, 391, (1905). 
 
 2 McCoy and Goettsch, loc cit. 
 
10 
 
 The analysis of thorium minerals. 
 
 The activities of several thorium minerals were determined by 
 the first method described in the same way as for sample A of pure 
 thorium dioxide. The ore was ground up very finely in alcohol, and 
 care was taken to prevent any segregation of the more active consti- 
 tuents of the ore in the preparation of the films. Sampled portions 
 of the powdered ore were then taken for the purpose of determining 
 the percentage of thorium and uranium present. The uranium was 
 determined by reduction with zince in sulphuric acid and titration 
 with potassium permanganate. 1 The analyses for thorium were 
 made by the method of Neish. 2 Instead of evaporatin to dryness 
 the solution obtained by dissolving the precipitated hydroxides in 
 dilute nitric acid, this solution w r as almost neutralized by adding 
 dilute ammonia, methyl orange being used as indicator. 250 CC of a 
 solution of meta-nitrobenzoic acid, made by dissolving 4 grams of 
 the acid in a liter of water, was then added slowly with constant 
 stirring. After being warmed to 80 for a time, the precipitate , 
 which collected and settled to the bottom, was filtered off and washed 
 with a 5 per cent, solution of the precipitant. The precipitate was 
 then dissolved in dilute nitric acid, diluted, to about 150 CC , almost 
 neutralized with dilute ammonia as before, and the thorium then 
 leprecipitated by the addition of 200 CC more of the nitrobenzoic acid 
 solution. By this procedure, which shortened somewhat the method 
 outlined by Neish, pure white thorium dioxide, giving very con- 
 cordant results, was always obtained. 
 
 The locality of origin, and the percentage of thorium and uran- 
 ium in the minerals investigated are shown in Table III. 
 
 TABLE III. Thorium Minerals. 
 
 Per cent. 
 No. Mineral. Thorium. Uranium. 
 
 1. Orangite, Langesundsfjord, Norway 43*1 7*76 
 
 2. Thorite " ' 46-6 6-26 
 
 3. Monazite, McDowell Co., N. C 5-27 -33 
 
 4. " Roada, Norway , 15*18 -46 
 
 5. " Commercial sample. . 2*7 2 -12 
 
 1 The details of the method of separation and determination of uranium are 
 given in Part II. 
 
 2 Loc, cit. 
 
11 
 
 The specific activity of thorium minerals. 
 
 The activity of films of thorium minerals was corrected for the 
 effect of the emanation according to the formula applied to films 
 of pure thorium dioxide. In calculating a from & 1? the approxi- 
 mate value of ki, as found from the uncorrected activity of the 
 thinner films, was used. This procedure involves no appreciable 
 error in the calculation of the true activity. 
 
 The results are given in Tables IV to VIII. The significance of 
 the symbols is the same as given under sample A. The curves for 
 all the minerals were smooth. It is not necessary to reproduce them 
 here. 
 
 TABLE IV. Orangite, No. 1. 
 
 w 
 
 1-2400 
 3415 
 2259 
 1212 
 0526 
 
 52-29 
 45-01 
 39-24 
 
 28-46 
 14-85 
 
 47-2 
 44-1 
 38-8 
 28-3 
 14-8 
 
 (a- 
 
 w 
 
 1825 
 3243 
 2339 
 1349 
 0904 
 
 =00310 
 
 TABLE V. 
 a, 
 
 50-51 
 43-03 
 39-37 
 29-74 
 22-61 
 
 w 
 
 a 
 
 02628 
 00775 
 00583 
 00428 
 00355 
 
 G): 
 
 -22 
 
 Thorite, 
 
 No. 2. 
 
 
 w 
 
 a 
 
 
 
 
 a 
 
 47-1 
 
 02511 
 
 42-4 
 
 00765 
 
 39-0 
 
 00600 
 
 29-6 
 
 00456 
 
 22-6 
 
 00400 
 
 (a- 
 
 =00301 
 
 6). 
 
 26 
 
 TABLE VI. Monazite, No. 
 
 w 
 
 1-0587 
 
 2825 
 2059 
 1094 
 0476 
 
 o, 
 
 4-70 
 3-96 
 3-43 
 2-36 
 1-23 
 
 - 1 =-0332 
 
 4-55 
 3-94 
 3-42 
 2-35 
 1-23 
 
 a 
 
 2330 
 0717 
 0602 
 0466 
 0387 
 
 w\ 
 
 - I =6-2 
 
 E 
 
 5-80 
 
 1-50 
 
 1-00 
 
 (55) 
 
 (24) 
 
 E 
 
 3-90 
 1-16 
 
 88 
 (-53) 
 (.34) 
 
 E 
 
 .17 
 
 .04 
 
 .03 
 
 (.02) 
 
 (.01) 
 
12 
 
 TABLE VII. Monazite, No. 4. 
 
 w 
 
 w j a E 
 
 a 
 
 1.1081 12-83 12-27 0903 -64 
 
 2973 10-75 10-61 0279 -17 
 
 2123 9-53 9-48 -0224 -12 
 
 1078 6-L'6 6-24 -0173 (-06) 
 
 0526 3-68 3-67 -0143 (-03) 
 
 w\ 
 
 )=' 0122 
 
 TABLE 
 
 VIII. Monazite, 
 
 J^o. 5. 
 
 
 
 w 
 
 <Zi 
 
 a 
 
 
 
 
 
 a 
 
 2-15 
 
 
 493 
 
 1-89 
 
 >> 
 
 168 
 
 1-70 
 
 Is 
 
 A, 
 
 127 
 
 1-37 
 
 ' * 
 
 100 
 
 1-25 
 
 1 s 
 
 094 
 
 89 
 
 ^"^ 
 
 088 
 
 46 
 
 1 1 ' 
 
 076 
 
 
 ( - ) =-066 
 \a / 
 
 
 w a a E 
 
 a 
 
 1-0590 2-15 -493 J 
 
 3174 1-89 >? -168 
 
 2156 ' 
 
 1375 
 1170 
 0785 
 0348 
 
 -o 
 
 In Table IX, & lw represents the specific activity of the mineral 
 
 (w\ 
 
 as calculated from k lm =2/i - I . k w represents the activity due to 
 
 uranium and its products contained in 1 gram of the mineral. This 
 is found by multiqlying the weight of uranium in 1 gram of the 
 mineral by 3280; since, as has been mentioned, the specific activity 
 of a thorium free mineral containing 1 gram of uranium is 3280 
 units. ki Th is the difference between k lm and k lff ; it represents the 
 activity due to thorium and its products in 1 gram of the mineral. 
 In the last column P rfc is the weight of thorium in 1 gram of the 
 
 mineral. -^-is as nearly constant as one could reasonably 
 
 *Th 
 
 expect, considering that all the errors of experiment are accumu- 
 lated on this ratio. It may be concluded, therefore, that the specific 
 radioactivity due to 1 gram of thorium in any mineral is a constant, 
 and is equal to 953 units. The radioactivity of any mineral con- 
 taining uranium and thorium which is sufficiently old geologically, 
 is therefore equal to 3280 P v + 953 P Th . 
 
13 
 
 TABLE IX. Thorium Minerals. 
 No. Name. %Th %U k lm k w k lTh |^ A 
 
 * Th 
 
 1. Orangite 43-10 7.76 649 255 394 914 
 
 2. Thorite 46-60 6-26 664 2U5 459 985 
 
 3. Monazite 5^7 -33 60-2 10-9 49-3 935 
 
 4. " 15-18 -46 164 15 149 982 
 
 5. 272 -12 29-8 4-0 25-8 950 
 
 Mean =953 
 
 The fact that the specific activity of thorium in minerals is found 
 to be constant makes it necessary to assume that the .activity of thor- 
 ium is not due to radioactive bodies accidentally retained by the 
 thorium (as radium frequently is by barium sulphate), but that 
 radiothorium is a disintegration product of ordinary thorium with 
 a relatively slow period. 
 
 The specific activity of thorium in thorium dioxide recently 
 extracted from thorium minerals. 
 
 In the analyses of the minerals, the thorium was weighed as 
 dioxide. This was very pure chemically. During the course of the 
 analysis, the thorium was precipitated once by oxalic acid, once by 
 potassium hydroxide, and twice by meta-nitrobenzoic acid. The 
 thorium residues from each of the minerals were kept separate, and 
 made into films. Measurements were made of the activities of the 
 films one month later. The thorium dioxide then contained the 
 maximum amounts of Th X, Emanation, Th A and Th B. The 
 results are given in Tables X to XII, which represent the samples 
 of dioxide prepared from the minerals, No. 1, No. 2 and No. 4 
 respectively. A summary is given in Table XIII. 
 
 TABLE X. Thorium dioxide B. 
 
 w 
 
 w di a E 
 
 a 
 
 2267 61-4 60-9 -00372 1-04 
 
 1170 40-2 40-1 -00292 (-50) 
 
 0673 25-1 -25-1 -00268 (-30) 
 
 = 00243 
 
14 
 
 TABLE XI Thorium dioxide C. 
 
 w 
 
 2634 
 1292 
 0657 
 
 68-2 
 41-8 
 24-i 
 
 67-8 
 41-7 
 24-1 
 
 - \ =-00230 
 
 a 
 
 00389 
 00510 
 00273 
 
 E 
 
 1-04 
 
 (-50) 
 
 (20) 
 
 w 
 
 2088 
 1083 
 0585 
 
 TABLE XII. Thorium dioxide D. 
 
 57-6 
 35-9 
 21-6 
 
 57-2 
 35-7 
 21 5 
 
 (D- 
 
 002 *0 
 
 a 
 
 00365 
 00303 
 00272 
 
 E 
 
 1-10 
 
 (60) 
 
 (-30) 
 
 TABLE XIII. Thorium dioxide. 
 
 Sample 
 
 Source 
 
 C). 
 
 A Thorium Nitrate -00387 
 
 B Orangite No. 1 . '00243 
 
 C Thorite No. 2 -00230 
 
 D Monazite No. 4.. . -00240 
 
 517 
 
 823 
 870 
 833 
 
 P 
 
 588 
 936 
 989 
 948 
 
 Mean of B, C and D = 958 
 
 P represents the weight of thorium in 1 gram of the oxide. The 
 
 k 
 
 mean value of ^ for the last three samples is 958, which represents 
 
 the total activity of 1 gram of thorium in the oxide. The corres- 
 ponding value fbr thorium in the minerals is 953. Since the activ- 
 ity of thorium in samples B, C and D, and in minerals is practically 
 identical, it follows that no appreciable fraction of the radiothorium 
 present in the minerals has been removed by the processes of analy- 
 sis. The oxides obtained from the commercial thorium .salts, e. g. } 
 sample A, are however only a little more than one-half as active as 
 the oxides contained in, or separated from, the natural thorium 
 minerals. 
 
15 
 
 On the separation of radiothorium from thorium. 
 
 The least active thorium compound described, sample A, gave 
 off an appreciable amount of emanation. It must, therefore, have 
 contained a considerable portion of the equilibrium amount of radio- 
 thoriura, which may be symbolized Bit. I have attempted by a 
 number of methods to remove the radiothorium completely and so 
 obtain thorium o'f minimum activity. The experiments carried out 
 and the results obtained may now be described. 
 
 It was first shown by Rutherford and Soddy 1 that Th X and 
 subsequent products may be removed from a compound of thorium 
 in solution by repeating a few times the precipitation of the thorium 
 by ammonia. By this treatment the Th X passes into the filtrate, 
 and the activity of the thorium compound is thereby greatly reduced. 
 In the course of five or six weeks, however, in Rutherford and 
 Soddy's experiments, the activity seemed to be completely regained, 
 owing to the reproduction of Th X, etc. To ascertain if any of the 
 radiothorium was removed by precipitation with ammonia, a quan- 
 tity of thorium nitrate from the same lot from which sample A had 
 been made, was precipitated with ammonia. The precipitated 
 hydroxide was filtered off, dissolved in dilute nitric acid, and thie 
 process repeated 100 times. The last precipitate was ignited; to 
 thorium dioxide, and its activity determined immediately and at 
 intervals throughout a period of several weeks. The specific activ- 
 ity at first showed a minimum value of ^=185. The activity 
 increased and reached a maximum at the end of five weeks when 
 &j = 453. The result shows that some of the radiothorium had been 
 removed, since the value of ki found for sample A was 513. That 
 a considerable amount of radiothorium still remained, however, was 
 shown by the fact that the minimum activity increased at a rate 
 which indicated the production of Th X and subsequent products. 
 
 Another portion of the nitrate solution containing 10 grams of 
 salt was treated with 40 CC of a 3 per cent, solution of hydrogen per- 
 oxide. The precipitated thorium was separated from the filtrate 
 by a Pukall filter. The precipitate was dissolved in 15 CC of concen- 
 trated nitric acid, diluted and partly neutralized with ammonia, 
 and the thorium again precipitated by adding the same amount of 
 hydrogen peroxide. This process' was repeated 18 times. The final 
 
 U'roc. Chem. Soc., 18, 120, (1002). 
 
16 
 
 precipitate, which was reduced to a small amount through loss in 
 precipitation , was carefully purified by Neish's method, and then 
 ignited. Th X and subsequent products were removed by this 
 treatment. The Initial specific activity was 163, and the maximum, 
 reached after about four weeks, was 377. It follows, therefore, that 
 18 precipitations with hydrogen peroxide do not completely remove 
 radiothorium although the maximum activity has certainly been 
 considerably reduced. 
 
 The thorium in a third portion of the nitrate solution was pre- 
 cipitated as thorium acetylacetonate by adding a slight excess of 
 sodium acetylacetonate. The precipitate was washed with alcohol, 
 dried, and then sublimed in a vacuum of 1 mm. of mercury. Por- 
 tions of t'he sublimed salt from different parts of the tube were 
 separately collected, and their activities measured. Three portions 
 showed minimum specific activities of 167, 192 and 188, and maxi- 
 mum activities, one month later, of 444, 488 and 460 respectively. 
 
 A solution of thorium nitrate was precipitated fractionally with 
 a solution of oxalic acid. The first of the eight fractions had a mini- 
 mum activity of 208, and a maximum activity one month later of 
 476. The eighth fraction showed the same maximum activity. 
 
 In addition to these reactions, the thorium was precipitated, 
 many times repeated, with such reagents as, potassium chromate, 
 sodium thiosulphate, sulphuric acid in rather concentrated solution, 
 etc. In a solution of thorium nitrate was precipitated repeatedly 
 small amounts of barium sulphate by adding alternately barium 
 chloride and sulphuric acid. Results differing but slightly from 
 those given above were obtained in each case. 
 
 Some of the more important results are collected in Table XIV. 
 
 TABLE XIV. 
 
 mm. max. max. 
 
 100 precipitations with ammonia 185 435 2.35 
 
 18 hydrogen peroxide .... 163 377 2.31 
 
 Sublimation of the acetylacetonate 167 444 2.66 
 
 Oxalic acid precipitation 1st fraction 208 476 2.29 
 
 8th " 476 
 
 8 precipitations with sodium thiosulphate 190 482 2.54 
 
 7 " potassium chromate 196 494 2.52 
 
 Mean 2.45 
 
17 
 
 While the experiments did not lead to the desired result the 
 complete separation of radio>thorium from thorium they serve to 
 emphasize the remarkable persistency with which radiothorium is 
 retained by thorium. It seems to be a very easy matter to separate 
 Th X and subsequent products, as shown by the experiments 
 described., as well as by the earlier work of Moore and Schlundt. 1 
 
 Two new papers by Hahn 2 are important in connection with the 
 foregoing. Hahn finds that freshly made commercial samples of 
 thorium are more active than older ones. The activity was smallest 
 for samples three to nine years old. A sample twelve years old had 
 greater activity. It is suggested in explanation that radiothorium 
 has a period of about two years (a similar paper by Blanc 3 gives 
 the period as 737 days), but that it is formed from an inactive pro- 
 duct which is in turn the direct product of thorium. This inter- 
 mediate product, called mezothorium, was isolated in an impure 
 form by Hahn. It gives thorium emanation, and its activity increa- 
 ses steadily with time. The spontaneous loss of activity of commer- 
 cial thorium is explained as due to the more or less complete removal 
 of mezothorium in the process of preparation. Radiothorium and 
 subsequent products then decay with time. 
 
 In Table XIV, it appears that the activity after the maximum 
 amount of Th X has formed is in exery case appreciably leiss than 
 the activity of sample A for which &, =513. The small activity 
 represents a smaller proportion of radiothorium and products than 
 was present in sample A. The portion of the radiothorium lost 
 may have been removed chemically., or it may have simply decayed 
 in the time which elapsed between the preparation and measurement 
 of sample A and the samples in question. An approximate calcula- 
 tion based on Blanc's value for the period of radiothorium indi- 
 cates that all cases, excepting where the purification was by means 
 of hydrogen peroxide, natural decay alone of radio>thorium would 
 account very largely for the dimunition of activity observed. 
 
 I have measured the activity of some samples of thorium dioxide 
 which had been prepared and purified by Neish's method one and 
 one-half years ago. The activity had decreased about 30 per cent. 
 The decrease is less than that calculated by Blanc's value for the 
 
 1 Jour. Phys. Chem., 9, 682, (1905). 
 
 2 Ber. d. chem. G*s.. 40. 1462 and 3304, (1907). 
 3 Physik, Z., 8, 321, (1907). 
 
18 
 
 decay constant of radio thorium,, assuming the total activity to be 
 due to radiothorium and subsequent products, and that the rate of 
 reformation of radiothorium is negligible. These results indicate 
 that thorium alone is active. The activity of one gram of thorium 
 freed from all its products is probably between 100 and 130. Fur- 
 ther detailed experiments are necessary to determine with accuracy 
 the activity of thorium when free from all its products. 
 
 PART IT. 
 
 Uranium Compounds. 
 
 As has been already mentioned in connection with the thorium 
 compounds, the investigations of McCoy 1 and Goettsch 2 showed 
 that the total activity of one gram of uranium is independent of 
 its form of chemical combination and is equal to 790 units, the 
 unit being the observed activity of 1 sq. cm. of a thick film of the 
 oxide, UgOg. The total activity associated with one gram of uran- 
 ium in a mineral was found to be 3280 units. Minerals were there- 
 fore found to be 4.15 times. as .active as pure compounds of equal 
 uranium content. Rutherford showed that the maximum ioniza- 
 tion of a given substance is not produced if the distance between 
 the electrodes is too small. This fact was completely explained by 
 the work of Bragg and Kleeman, 3 who showed that the a-rays of 
 any radioactive substance have, in air, a definite range throughout 
 which they produce ioniza-tion. Maximum ionization is produced 
 only when the ionization chamber is sufficiently large to permit all 
 the a-rays to reach the limits of their ranges. The ranges of radium 
 and its products 4 are between 3.50 end 7.06 cm., while that of 
 uranium is probably near 3.5 cm. 5 The electroscope used by McCoy 
 had the electrodes set at 3.5 to 4.5 cm. While these distances are 
 doubtless great enough for pure uranium compounds, they are too 
 small for the minerals the activity of which is due, in large mea- 
 sure, to radium and its products. The new experimental results, 
 to be described, justify this conclusion. The activity of uranium 
 in pure compounds was practically unchanged when measured with 
 
 1 Loc. cit. 
 
 2 Loc. cit. 
 
 3 Phil. Mag., 10, 318 and 600, (1905). 
 
 4 See also Bragg, Phil Mag., 8. 726, (1904). 
 
 5 Bragg, Phil. Mag., 11, 754, (1906). 
 
19 
 
 electrodes at a distances of 8.5 cm., while under the new conditions 
 of measurement, the activity of a mineral was about 4.0 per cent, 
 greater. 
 
 The recent work on the range of the a -rays also served to indi- 
 cate another possible error in the method of McCoy. According to 
 this method, films of the active material were deposited in flat, cir- 
 cular metal dishes 7.02 cm. in diameter with rims 0.8 cm. high. 1 
 Inasmuch as the solid angle subtended by the rim, with any point 
 of the film as centre, was the same for all films, the rims being of 
 equal height, it was thought at first that the same fraction of the 
 radiation would be cut off by the rims in all cases. 2 However, 
 films sending out a -rays with greatest effective range will suffer 
 greatest loss of activity on account of the rims. Experiment con- 
 firmed this supposition, although the difference in percentage loss 
 was not great. Thus, while the actual dimunition of the activ- 
 ity due to the rims was 8.2 per cent, for uranium oxide, the corres- 
 ponding decrease in activity in the case of uranium minerals 
 amounted to 8.8 per cent., a difference of 0.6 per cent. 
 
 Finally, in order to obtain exact results, the observed activity 
 of a film of a uranium ore must be corrected for the activity of the 
 emanation which is lost spontaneously. Boltwood 3 has determined 
 the percentage of emanation thus lost for a large number of uran- 
 ium minerals and has found appreciable loss. in every case. 
 
 Experimental part, the effect of insufficient ionization space. 
 
 In order to obtain results free from the above mentioned errors, 
 a new series of determinations was undertaken. A new electroscope 
 was constructed for the activity measurements. It was made of 
 sheet brass, 1.4 mm. thick, and consisted of two superimposed rec- 
 tangular compartments., a smaller upper one 7.5 x 10 cm. and 12.5 
 high, which contained the gold leaf system, and a larger lower one 
 19.5 cm. square and 14 cm. high, which, constituted the ionization 
 chamber. The circular brass plate which served as the upper elec- 
 trode had a diameter of 14 cm. The film, whose activity was to be 
 measured, was placed on a movable support in the lower part of the 
 ionization chamber and so arranged that the distance between the 
 film and the charged electrode could be changed at will. The gold 
 
 1 Tin plate jelly-glass covers were used. 
 2 Goettsch, J. Am. Chem. Soc., 28, 1548, (1906). 
 3 Phil. Mag., [6], 9, 599, (1905). 
 
20 
 
 leaf-system was connected with the electrode by a brass rod which 
 passed through an insulating amber support. The motion of t<he 
 gold leaf was observed by a micrometer microscope, the upper com- 
 partment being provided with glass windows. The whole instru- 
 ment was surrounded by a wooden case to guard against any sudden 
 changes in the temuerature, or effect of air currents. To charge the 
 electroscope a difference of potential of about 450 volts was required. 
 
 In order to determine the effect due to insufficient ionization 
 space in the electroscope previously used, a series of films of Pitdh- 
 blende No. 1 were prepared in the same way as described for sample 
 A of thorium dioxide. The activity of the films was then mea- 
 sured in the electroscope with the charged plate placed first at a 
 distance of 4.5 cm. from the active film, and then at a distance of 
 8.5 cm. With the plate placed at the latter distance, the ionization 
 space in the electroscope is sufficient to permit all the a-rays given 
 off by the mineral to reach the end of their free paths before being 
 stopped by the charged plate, or by the walls of the electroscope. 
 For these measurements the films were deposited as usual in tin 
 dishes with rims. These dishes were 7.02 cm. in diameter, and 0.8 
 cm. high. On making the two sets of measurements, and calculat- 
 ing the specific activity of the ore from the results obtained, it was 
 found that with the plates only 4.5 cm. apart the activity indicated 
 was 4.0 per cent, too low. Accordingly all subsequent measurements 
 were made with a distance between the plates of 8.5 cm. 
 
 On the effect due to the rims.. 
 
 To determine the effect due to the rims, a series of films of pure 
 uranium oxide were made in the ordinary dishes with rims 0.8 cm. 
 high and their activities measured. The rims were then carefully 
 cut off with a pair of snips. The thick film used as standard was 
 likewise measured first with the rim on, and then after it had been 
 cut off. It was found that the specific activity of the oxide, as cal- 
 culated from the measurements made with the rims on, was 8.2 per 
 cent, less than the corresponding value obtained with the rims off. 
 
 Films were also made of three uranium ores in the .same way as 
 for the pure oxide. The ores were carefully sampled, and special 
 care was taken when preparing the films to prevent any segregation 
 of the more active constituents. Activity measurements were made 
 first with the rims on, and then after they had been cut off. It was 
 
found, as in the case of the pure oxide, that the specific activity of 
 each of the numerals,, as calculated from the measurements made <-f 
 films with rims, was less than the corresponding value obtained 
 when the films were measured without rims. The average decrease 
 in activity due to the rims for the three minerals amounted to 8.8 
 per cent. The corresponding value obtained for the pure oxide was 
 8.2 per cent. 
 
 All films subsequently prepared were therefore deposited on per- 
 fectly flat surfaces. This was most conveniently done by placing 
 false bottoms made of sheet copper in the vessels previously used. 
 The proper amount of the finely powdered material to be used in 
 making a film wias then vigorously shaken up with alcohol in a small 
 beaker and quickly poured into one of the dishes in the usual way. 
 When the alcohol had evaporated, the material was left deposited 
 on the false copper bottom which could then be removed, weighed, 
 and its activity determined. The diameter of each copper plate was 
 7.00 cm. 
 
 The specific activity of uranium in uranium oxide. 
 
 In Table XV, is given the results obtained when a series of 
 films of pure uranium oxide was made on flat copper plates in the 
 way described. The standard of activity used was a thick film of 
 the oxide weighing about 0.025 gram per sq. cm., and deposited on 
 a copper plate of the same area as the others. The activity of 1 
 sq. cm. of such a film, was taken as unit activity. The significance 
 of the symbols is the same as already given in the description of the 
 thorium compounds. 
 
 TABLE XV. Uranium oxide, U s O g 
 
 w 
 
 w a 
 
 a 
 
 .6040 38.7 .01560 
 
 2902 36.5 .00795 
 
 .2054 33.1 .00621 
 
 ,1048 23.0 .00455 
 
 0757 18.7 00404 
 
 ,0534 14.6 .00365 
 
 ( w }= 
 
 \ a /* 
 
 = 00296 
 
22 
 
 /w\ 
 Since &i=2/I - 1 the true specific activity of uranium oxide 
 
 in terms of the activity of 1 sq. cm. of a thick film of the oxide is 
 2 
 
 - = 676. Uranium oxide contains 84.85 per cent, of uran- 
 
 ium. Therefore, the true total activity of 1 gram of uranium in 
 
 676 
 uranium oxide is = 796 units. 
 
 Uranium ores, Method of analysis. 
 
 A series of experiments was likewise undertaken to determine 
 the total activity of uranium in its ores. The ores investigated were 
 powdered and then ground up finely in an agate mortar with alcohol 
 to prevent the possible loss of any highly radioactive dust. The 
 ground ore was then thoroughly mixed by sampling, and analyses 
 made for uranium and thorium. The same sampled portion of ore 
 was likewise used in making films, and in determining the amount 
 of emanation spontaneously evolved. 
 
 The method adopted in determining the percentage of uranium 
 in its ores is a modification of that of Brearly. 1 The method' of 
 procedure is essentially the same as that employed by Boltwood, 2 
 but, as no satisfactory results could be obtained when using most 
 of the reagents in the proportions which he directed, a number of 
 experiments were undertaken to determine to what extent the 
 method should be modified. Satisfactory results were finally 
 obtained when the analyses were carried out as follows : About 
 .3 gram of the finely ground ore was dissolved in dilute hyojro- 
 chloric acid with a little nitric acid added ; the solution was evapo- 
 rated to dryness, moistened with a little concentrated hydrchloric 
 acid and again evaporated, this being repeated three times to render 
 the silica insoluble. The solution was then taken up in dilute 
 hydrochloric acid and the insoluble residue filtered off. Hydrogen 
 sulphide was passed into the filtrate for an hour, the precipitated 
 sulphides filtered off, and after boiling to remove the excess of 
 hydrogen sulphide, the salts in solution were oxidized by adding a 
 little bromine water. On boiling a short time, a little more bro- 
 mine water was added, and the boiling continued until the excess 
 of bromine was driven off. The solution was allowed to cool some- 
 
 1 Loc. cit. 
 
 2 Phil. Mag., [6], 9, 603, (1905). 
 
23 
 
 what, .2 gram of microcosmic salt added, and then poured into a 
 solution of 30 grams of sodium carbonate (crystals) in 125 cc. of 
 water. The mixed solutions, which should not have a volume exceed- 
 ing 300 CC , was boiled gently for five minutes, allowed to cool some- 
 what, 20 grams of ammonium chloride added, and the boiling 
 repeated as before for five minutes. After 'allowing to stand for 
 some hours, the clear solution was decanted off, and the precipitate 
 washed. The filtrate was boiled in a large beaker until the ammon- 
 ium carbonate was driven off, and the uranium phosphate precipi- 
 tated from the solution. 10 cc. of dilute nitric acid was added to 
 the hot solution and the boiling continued until the uranium phos- 
 phate again dissolved. In the case of most ores, an insoluble resi- 
 due was found to remain after the uranium had all gone into 
 solution. This was filtered off, and the filtrate evaporated to 200 CC . 
 After allowing to cool, .8 gram microcosmic salt was added, and* 
 then 10 grams' of crystallized sodium thiosulphate. The solution 
 was heated very carefully to boiling, and boiled gently for 
 15 minutes, allowed to stand for a short time, the precipitated 
 sulphur 'and uranium phosphate then filtered through a platinum 
 Gooch crucible, washed, dried for an hour in the air bath, and then 
 ignited to burn off the sulphur. The residue was dissolved in as 
 small amount as possible of concentrated-nitric acid, the asbestos 
 filtered off, and ammonia added until a precipitate began to form. 
 The solution was then heated on the water bath and just enough 
 nitric acid added until the solution became clear again. After 
 allowing to cool, it was poured into a concentrated solution of 
 25 grams of ammonium carbonate in 100 cc. of water. Hydrogen 
 sulphide was passed in for 15 minutes, and, after heating to boil- 
 ing, it was allowed to stand for a few hours. The slight precipitate 
 formed was filtered off, the filtrate boiled to drive off excess of 
 ammonium corbonate, 10 cc. of dilute nitric acid added, and then, 
 on allowing to cool, .8 gram of microcosmic salt and 10 grams of 
 sodium thiosulphate. The solution was boiled for 15 minutes as 
 before, filtered through a Gooch crucible, washed, dried for an hour 
 in the air bath at 110, the sulphur carefully burned off and th|e 
 residue then heated to low redness for 15 minutes. The uranium 
 may be determined from the weight of the phosphate thus obtained. 
 It was found more satisfactory, however, to dissolve the residue in 
 lOOoc. of dilute sulphuric acid, reduce with zinc, and then titrate 
 
24 
 
 with a standard solution of potassium permanganate. 1 The per- 
 manganate solution was standarized first by means of pure iron 
 wire, and then by titrating a weighed amount of uranium oxide, 
 U 3 O 8 , dissolved in sulphuric acid and reduced with zinc. Both 
 results were found to agree exactly. 
 
 Table XVI gives the locality of origin, and percentage of uran- 
 ium in the three minerals analysed. All were analysed for thorium 
 by NeishV method, 5 gram portions of each ore being taken for 
 analysis, but no trace of thorium was found in any case. Since 
 it has been shown that the ratio of the activity of a mineral to its 
 uranium content is constant, it was not considered necessary to 
 investigate any large number of minerals in order to determine the 
 true value of this ratio. 
 
 TABLE XVI. 
 
 Per cent, of 
 Mineral. Locality. Uranium. 
 
 Pitchblende, No. 1 Unknown 44.7 
 
 No. 2 Central City, Col 54.0 
 
 No. 3 Gilpin Co./ Col 58.1 
 
 The observed specific activity of the minerals. 
 A series of films was made of each mineral in the same wiay as 
 for the pure uranium oxide. In the last column of Table XVII is 
 given the observed specific activity for each mineral. 
 
 TABLE XVII. 
 
 Mineral. 
 
 Of). 
 
 (uncorrected) 
 
 Pitchblende, No. 1 . . 001264 1582 
 
 No. 2 001048 1908 
 
 No. 3 000991 2018 
 
 Determination of evolved emanation. 
 
 In order to determine the true specific activity of the minerals 
 and the correct value of the ratio of the activity of the minerals to 
 that of the pure oxide containing the same amount of uranium, the 
 observed activity of each mineral must be corrected for the loss of 
 
 JGoettsch, J. Am. Chem. Soc., 28, 1547, (1936). 
 2 Loo. cit. 
 
25 
 
 the evolved emanation, as was already stated. The emanating power 
 of each mineral was determined .separately. For this purpose a por- 
 tion of the ore, ground to the same degree of fineness as that subse- 
 quently used in determining the retained emanation, was carefully 
 weighed and introduced between plugs of cotton wool in a straight 
 calcium chloride tube. A slow current of air was then drawn! through 
 the tube for 10 minutes to remove any emanation which might be 
 clinging to the solid particles. The tube was then tightly stoppered, 
 and allowed to stand for a month. At the end of this time, the 
 evolved emanation which had collected was removed by drawing 
 slowly through the tube about 250 cc. of air. This was collected in 
 a flask, allowed to stand 15 minutes, and then introduced into the 
 gas electroscope which had previously been exhausted by a water 
 pump. At the end of two hours, it was found that the activity of 
 the gas in the electroscope had reached a value remaining constant 
 for a considerable length of time. Accordingly, in every case, the 
 emanation was allowed to stand in the electroscope two hours before 
 noting its activity. The observed activity, divided by the weight 
 of the ore taken, gives the activity of the emanation lost by one 
 gram of the mineral. 
 
 Determination of retained emanation. 
 
 To determine the activity of the emanation retained by the ore, 
 a weighed portion,, about .03 gram, was placed in a small flask, 
 treated with a mixed solution of 20 cc. dilute nitric acid and 5 cc. of 
 dilute sulphuric acid. The small flask was connected with a liter 
 flask filled almost to the neck with water to which was added a few 
 cc. of a sodium hydroxide solution. This flask was in turn con- 
 nected by means of a siphon with a third vessel. On heating the 
 solution in the small flask to boiling,, the ore was decomposed, and 
 the gases and steam carried the emanation over into the liter flask, 
 displacing the water. The boiling was continued 10 minutes. The 
 liter flask containing the solution was then tightly stoppered and 
 allowed to stand 15 minutes to permit of the decay of the actinum 
 emanation. The inclosed gas was then introduced into the electro- 
 scope as before, and the activity noted at the end of two hours. The 
 activity due to the emanation retained in 1 gram of the ore, as thus 
 determined, plus the activity of the evolved emanation as previously 
 found, gives the total activity of the emanation in 1 gram of the 
 substance. 
 
The following Table gives the values found, and the per cent, of 
 the emanation spontaneously lost by each mineral. The activities 
 are given in terms of the fall of the leaf in scale divisions per 
 second. Each number represents the mean of three determinations 
 all of which agreed closely. The values in the second, third and 
 fourth columns refer to 1 gram of the ore. 
 
 TABLE XVIII. 
 
 *oc *o c "o c a * A^ 
 
 ^ o ^ tek S a * 2 .0 sS ^1 
 
 ^j F 8 '43 +j T3 '-M Jj -^ a --3 'S T5 - 1 ao a 
 
 Mineral. ^| g j> g ;g ^ ^ 8 f IJ g- g" 7 J 
 
 Pitchblende No. 1. -G29 '0246 -654 3'76 '447 1-46 
 
 No. 2 -0649 -786 8-25 -540 1-456 
 
 No. 3. -800 -0369 -887 4-41 -581 1-44 
 
 The last column of Table XVIII shows that the activity of the 
 total emanation is sitrictly proportional to the percentage of uran- 
 ium in the mineral. This is in agreement with the results found by 
 Boltwood. 1 
 
 When the amount of retained emanation in ore No. 2 was deter- 
 mined in the way describel, it was found that the amount obtained 
 was less than was expected when compared with the results obtained 
 for the other two ores. A comparatively large residue was left when 
 this ore was treated with the mixed solution of nitric and sulphuric 
 acids. A small portion, about .06 gram, was then weighed out in a 
 platinum crucible. About 4 grams of potassium acid sulphate, 
 which had been heated to quite fusion, was poured into the crucible 
 containing the ore, and the heating continued until the ore had 
 completely dissolved. On cooling, the fused mixture was dissolved 
 in acidified water, boiled for five minutes to drive off all emanation, 
 and then sealed up in a small flask. After standing for four days, 
 the emanation was again boiled off, collected in the usual way, and 
 its activity observed. The radium emanation reaches half its equili- 
 brium amount in four days. Hence, twice the activity observed, 
 divided by the weight of the ore taken, gives directly the activity 
 of the total emanation in 1 gram of the ore. The value thus 
 obtained, .786, must represent the true activity of the total emana- 
 tion for ore No. 2, since it was brought into complete solution by 
 the fusion with potassium acid sulphate. As the value obtained by 
 
 1 Phil. Mag., [6], 9. 599, (1905). 
 
treatment with the mixed acid solution was much less than this, 
 it follows that all the emanation can not always be expelled from an 
 ore by boiling with an acid solution only. 
 
 The percentage loss of activity of a uranium mineral due to tht, 
 evolved Emanation. 
 
 It is probable that the emanating power of a compact mineral 
 like Pitchblende is negligibly small until the mineral is powdered. 
 If so, the loss of activity must date from the time when the mineral 
 is pulverized. During a few weeks, or even months, after pulveriza- 
 tion, loss of emanation will therefore -cause an appreciable deficiency 
 in the amount of the products of rapid decay, namely, RaA, RaB 
 and RaC only; while the amount of RaD, which has a 
 period of about 40 years, will not be noticeably altered. 
 The amounts of RaE and RaF, which are products of RaD, 
 will also be unchanged by loss of emanation. In order that one may 
 know by what fraction the specific activity of a mineral is decreased 
 by the spontaneous loss of emanation after pulverization, it is neces- 
 sary to know what fraction of the total activity is due to the ema- 
 nation and the products RaA, RaB and RaC. The experimental 
 determination was made as follows : A portion of ore No. 1, finely 
 ground, was treated with nitric acid, evaporated to dryness, and the 
 process repeated three times at intervals of two hours to admit of 
 the decay of the radium products of short period. After the third 
 evaporation, the residue was heated just strongly enought to decom- 
 pose the nitrates present. The resulting oxides were then ground 
 up finely in alcohol. The alcohol was evaporated off on a water bath. 
 The fine powder remaining was made into films as quickly as pos- 
 sible, and their activities determined in the usual way. The mean 
 of two experiments gave for the specific activity of the ore, when 
 thus treated, a value of 51.5 per cent, less than the original value. 
 Since, by this treatment, all the emanation originally retained by 
 the ore, as well as the products of short period, was removed, it 
 follows that the activity due to the radium emanation, and its pro- 
 ducts RaA and RaC, is equal to approximately one-half of the total 
 activity of the ore. 
 
 The relative activity due to the emanation, RaA, RaB and .RaC, 
 may also be calculated from theoretical considerations. Boltwood 1 
 
 1 Am. J. Sci., 21, 413, (1906). 
 
28 
 
 has shown by direct experiment that the activity of RaEmf 
 ,RaA+RaB+R,aiC is 4.64 times as great as that of the equili- 
 brium amount of radium. His experiments also indicate, as he has 
 pointed out, that the activity of the equilibrium amount of any 
 member of a radioactive series is proportional to its range. The 
 range of the a-rays of RaF is 3.85 cm., while that of radium is 3.50 
 
 3-5 
 cm. Consequently, RaP is O~YQ =1.10 times as active as the 
 
 equilibrium amount of radium. Since llaD is inactive, and RaE 
 gives only yS-rays, the activity due to the Em + RaA + RaB + RaC = 
 
 4. 64 
 
 -=- times that of radium and all its products. Now, uranium 
 b * 74 
 
 minerals are 4.54 times as active as pure compounds of equal uran- 
 ium content, as will be shown below. Therefore, radium and its 
 products are (4.54 1) =3.54 times as active as the equilibrium 
 amount of uranium. Consequently, the activity of RaEm+RaA+ 
 
 RaB + RaC = T~=~A = 0.536 of the total activity of a mineral. 
 
 6-74: x 4-o4 
 
 This amount, 53.6 per cent., agrees fairly well with that found by 
 direct experiment which gave 51.5 per cent. We may therefore con- 
 sider that the percentage loss of activity due to lost emanation of a 
 uranium mineral, which has been in the powdered state not more 
 than a few months, is equal to one-half of the emanation evolved. 
 
 The specific activity of uranium in minerals. 
 The proper corrections may now be applied to the observed spe- 
 cific activities of the ores as given in Table XVIII. Table XIX 
 gives a summary of the results obtained. 
 
 TABLE XIX. 
 
 Ore 
 
 
 it 
 
 ^3 
 
 l| 
 
 c'> 
 " 
 
 pq 
 
 ||| 
 fc rt 
 
 if 
 
 D , aT 
 
 fe 
 
 P 
 
 No. 
 
 1 
 
 1582 
 
 PL, 
 3 
 
 0> ( 
 
 .76 
 
 JH o 
 1 
 
 .9 
 
 r - 
 
 98.1 
 
 1613 
 
 .447 
 
 3610 
 
 No. 
 
 2 
 
 1908 
 
 8 
 
 .25 
 
 4 
 
 .1 
 
 95.9 
 
 1990 
 
 .540 
 
 3685 
 
 No. 
 
 3 
 
 2018 
 
 4 
 
 .41 
 
 2 
 
 .2 
 
 97.8 
 
 2064 
 
 .581 
 
 3552 
 
 Mean = 3616 
 
 The number 3616 represents the true specific activity of uran- 
 ium in its ores. The corresponding value for the activity of uran- 
 ium in its pure comjpounds is 796. Therefore, the true ratio of tdie 
 
29 
 
 activity of uranium minerals to the pure compounds containing the 
 
 . 3616 
 same amount ol uranium is _ = 4.54. 
 
 The specific activity of thorium in minerals. 
 The activity of the thorium films in the first part of this paper 
 was measured in an electrO'Scope having a distance between the 
 plates of only 4.5 cm. All measurements were made, moreover, 
 without the rims being removed from the tin dishes in which the 
 films were deposited. Two sources of error were thus introduced 
 in the calculation of the total activity of thorium, as the work on 
 uranium has shown. Since the percentage error due to these sources 
 is a constant one, or approximately so, this had no important effect 
 on the chief object of the research which showed that the specific 
 activity of thorium in its minerals is constant, while in its commer- 
 mercially prepared compounds it has a variable activity much less 
 than this. As the specific activity of thorium in its minerals is 
 constant, it was not considered necessary to repeat the measure- 
 ments of more than three of the minerals in order to determine the 
 true value of this constant. The large electroscope already referred 
 to in connection with the uranium minerals was used. The dis- 
 tance between the plates was 8.5 cm. The rims were removed from 
 the films before measuring their activities. Table XX which cor- 
 responds to Table XIII, gives a summary -of the results obtained. 
 The new value found for the total activity of 1 gram of uranium 
 in its ores 3616 units was used in calculating the activity due 
 to the uranium in each thorium mineral. 
 
 TABLE XX. Thorium minerals. 
 
 (w\ I- 
 
 -I k lm k w k lTh ,} Th 
 
 a /o / Tk 
 
 Orangite . . . 43.1 7.76 .00283 707 281 426 990 
 Thorite .... 46.6 6.26 .00286 699 226 473 1015 
 Monazite . 15.18 .46 .01160 172 16.6 155.4 1025 
 
 Mean = 1010 
 The mean of the numbers given in the last colum is 1010. This 
 
 represents the true total activity of 1 gram of thorium in its ores. 
 This investigation was carried out under the direction of 
 
 Professor H. N. McCoy, and I take much pleasure in expressing my 
 
 indebtedness to him for his many valuable suggestions and kind 
 
 supervision during the progress of the work. 
 
OVERDUE.