294 On the Relation between Uranium and Radium. 



others o£ similar character now in hand will throw further 

 light upon the early positive result obtained by the barium- 

 sulphate method of precipitation. But the following result 

 may be cited. The two fifty gram samples of uranyl nitrate, 

 one purified by barium sulphate and the other in the 

 condition purchased, except that it is dissolved in water, 

 which were used in the very first preliminary tests on April 1st 

 and March 31st, 1904, respectively, have recently been tested, 

 as before, by the bubbling method. They have been left in 

 their original bottles, the tubes closed with rubber caps, and 

 therefore no great precaution has been taken against the 

 possibility of contamination with extraneous radium. They 

 had not been opened for about three and a half years. 

 Initially the leak from the purified sample corresponded to 

 •013 of the present scale, and of the unpurified 0*4. Now 

 the quantity of radium in the two bottles is indistinguishable 

 by the rough bubbling test, and both give a leak by this test 

 of from 1 to 2. In the old kilogram of uranyl nitrate a leak of 

 about 15 was recorded after 550 days, corresponding to a 

 leak of 2 for 50 grams in 1580 days. These two samples 

 have now been sealed up in flasks and will be accurately 

 tested for amount of radium as soon as they are ready. Even 

 if they have been contaminated with radium, they will still 

 furnish a valuable maximum limit of the rate of production of 

 radium from commercial uranyl nitrate. The present rough 

 results show that this is extremely slow, and that the radium 

 present, with such small quantities of uranium, is even after 

 four years not at all conspicuous. Nevertheless it agrees 

 with the early result. 



Summary. 



(1) 250 grams of uranium (element) in the form of 

 uranyl nitrate, purified by extraction with ether and con- 

 taining initially 2' 3 X 10 -n gram of radium, has not increased 

 in radium in 600 days by more than 10 -n gram. 



(2) In another experiment with 481 grams of uranium 

 containing initially only 1*3 x 10 -11 gram of radium the 

 increase in amount of radium in 200 days is not greater than 

 5 x 10~ 12 gram. Both this and the preceding experiment 

 indicate that the production of radium is not greater than of 

 the order of one ten-thousandth part of the amount of uranium 

 disintegrating. 



(3) In 278 grams of uranium purified by the same process 

 there has been no appreciable growth of actinium in the course 

 of 300 days. 



