Rate at which Helium is Produced 379 Thus the difference between my results and Prof. Joly 's is somewhat emphasised .
I should have regarded my results , if they stood alone , as sufficiently conclusive , though it is impossible to help feeling disconcerted by want of agreement with so distinguished an experimenter .
It is perhaps possible , after all , that the difference is due to his having met with exceptional specimens .
Measurements of the Rate at Helium is Produced Thorianite and Pitchblendewith a Minimum Estimate of their Antiquity .
By the Hon. E. J. Strutt , F.E.S. , Professor of Physics , Imperial College of Science , South Kensington .
( Keceived July 23 , 1910 .
) S 1.\#151 ; Introductory .
The method of deducing a minor limit to the age of minerals from an examination of their radioactive properties has , up to the present time , depended on a measurement of the amount of helium they now contain , and on an indirect calculation of the rate at which it is being produced by the radioactive matter within them .
There is not now much uncertainty about this calculation .
Nevertheless , considering the fundamental importance of the question of geological time , it is not superfluous to determine in some favourable case by direct volume-measurement of the gas how much helium is produced per gramme of the mineral per annum , in order to see how long the quantity found in the natural mineral would take to accumulate , and to check the method of calculation to which we must still resort where the much more difficult direct method is impracticable .
A mineral suitable for such experiments must be obtainable by the kilogramme , and very radioactive , so as to give a measurable quantity of helium in a few months .
The minerals selected have been thorianite ( two varieties ) and pitchblende\#151 ; practically the only ones available .
Some account of preliminary work was given in a former paper.* Much more elaborate and satisfactory experiments have since been carried out .
These will now be described .
* ' Roy .
Soc. Proc. , ' A , vol. 83 , p. 98 .
380 Hon. R J. Strutt .
Rate at which Helium [ July 23 , The great difficulty of the problem lies in the small quantities of gas which have to be dealt with .
In this respect the conditions are far more onerous than in determinations of the helium production by radium ; for a quantity of radium equivalent in activity to a ton of pitchblende or thorianite may be placed in a small vessel , and the helium developed in a given time extracted and measured .
In the present investigation , the entire bulk of the original mineral has to be handled , and not merely the radium present in it .
It is impracticable to work with more than a few kilogrammes in this way , thus the quantity of helium which can be obtained in , say , six months is small indeed .
Assuming the difficulty of the measurement of small volumes got over , there remains another not less serious .
The helium initially present which has accumulated in geological epochs is perhaps 500 million times what the experimenter can grow under his observation .
Thus , to make the experiment satisfactory , it is necessary to remove the helium so perfectly that not more than one part in 5,000 millions of the original stock remains .
This could never be done if the mineral were allowed to remain in its original solid condition .
For solid minerals only yield their helium slowly and partially by heating.* It is essential , therefore , to get the mineral into solution , and to filter off any slight undissolved residue with the most scrupulous care .
Prolonged boiling will then remove helium with the necessary completeness .
It is important to use thick filter paper .
Fine particles of undecomposed mineral may otherwise get through the pores .
The importance of avoiding this will be understood when it is stated that the presence of 1/ 1000 of a milligramme of undecomposed thorianite was altogether inadmissible in my experiments .
Much trouble was incurred through a failure to adopt this precaution in the earlier attempts .
The necessity for dissolving the mineral raises a theoretical question of some importance .
Can we be really sure that this does not affect the rate of helium formation ?
I shall briefly discuss this question , chiefly for the convenience of those who approach the subject from a standpoint of general scientific interest , rather than as students of radioactivity .
Nearly all the evidence we have at the present time points to the conclusion that the rate of radioactive change is unalterable by anything that man can do .
It is true that a few experimenters have thought that they could detect changes in radioactivity at high temperatures , but the experiments of Bronson , f which have been pushed further than any others , reveal no such effect up to * Unless , indeed , very high temperatures are used , which would introduce many other difficulties .
t ' Boy .
Soc. Proc. , ' A , vol. 78 , p. 494 .
1910 .
] Produced in Thorianiteand , etc. 1600 ' C. The calorimetric experiments of Curie and Dewar showed no loss of heating effect in radium at liquid hydrogen temperatures .
As the development of heat is quantitatively accounted for by the expulsion of a-particles ( helium atoms ) with a high velocity , it cannot be doubted that the helium emission is unaltered at these temperatures .
More directly relevant are the experiments of Moore , * who found that radium emanation , dissolved in water , decayed at the same rate as when in the gaseous condition .
As the decay is the direct consequence of the emission of a-particles , it is clear that here also helium formation is independent of circumstances , and , in particular , of whether the radioactive body is in solution or not .
Lastly the method is only applicablef if we assume that the rate of helium production has been the same throughout the whole geological period which it is sought to measure .
What is there to be said in defence of this assumption ?
The critic will naturally object that the radioactive matter present is necessarily diminishing in quantity as it generates helium and other nonradioactive products to which it may give rise .
There must , therefore , have been more of it at the beginning of the geological period considered than at the end , and consequently more rapid production of helium .
The method as here applied is only valid if it can be shown that this diminution is unimportant in the period considered .
A simple argument goes far to establish this for thorianite .
This mineral is a dense substance , consisting almost entirely of the parent radioactive bodies , uranium oxide and thorium oxide .
We cannot suppose that there was ever much more of these bodies in the thorianite crystals than at present , for they do not contain much inactive matter of any kind , which can be assumed to represent the debris of the decayed radioactive bodies .
Apart from this simple argument , we have a good indirect estimate of the rate of decay of uranium , J which shows that it is unimportant in the periods here dealt with .
The decay of thorium is almost certainly slower still .
The other chief line of objection which can be taken is that the rate of formation of helium depends on the physical condition of the radioactive matter .
This objection has already been considered .
S 2.\#151 ; Experimental .
The thorianite used was dissolved in nitric acid , and most of the excess of acid driven off by evaporation .
A slight insoluble residue then remained , consisting in the main of zircon which had been mixed with the thorianite .
* ' Roy .
Soc. Proc. , ' A , vol. 80 , p. 597 .
t At least , in the simple form with which alone this paper deals .
f Boltwood , ' Amer .
Journ. Sei./ June , 1908 , vol. 25 , p. 506 .
VOL. LXXXIV.\#151 ; A. 2 D Hon. R. J. Strutt .
Rate which Helium is [ July 23 , This residue was filtered off and rejected .
It would no doubt have been preferable to reject nothing , but much additional trouble would have been incurred in getting it into solution .
The quantity of residue was insignificant and its activity much less than that of an equal weight of the original thorianite .
Two kinds of thorianite were used in the experiments , one the ordinary variety , the other that specially rich in uranium , from the Galle district .
Besides the experiments on thorianite some were made on pitchblende .
This mineral was treated with nitric acid , and the solution preserved separately .
Pitchblende contains sulphides which were oxidised by the nitric acid .
This led to precipitation of barium , radium , and lead as sulphates .
The insoluble residue was boiled with large excess of sodium carbonate to expel sulphuric acid , and , after washing , again treated with nitric acid .
After one or two repetitions of this process nothing but inactive sand remained .
This was rejected .
The solution obtained after the sodium carbonate treatment was , of course , kept separate from the original nitric acid solution .
The helium obtained from these separate solutions was all measured together .
The various solutions were stored in vacuous round-bottomed flasks arranged as shown at a in fig. 1 .
A tall tube b serving as a reflux condenser was sealed to the neck of the flask , and was continued in a narrow tube c , with recurved end dipping down into a basin of mercury .
The whole arrangement was fastened to a rod stand , and could be put away during the intervals allowed for the helium to be generated .
When it was desired to collect the gas the liquid was boiled , taking care to raise the temperature slowly so that no violent bumping occurred .
When vapour of appreciable pressure had formed over the liquid , boiling would proceed quite quietly , provided the solution was not too strong .
Boiling was continued by the application of a gentle heat , which did not raise more steam than could be condensed by the vertical tube , with simple air cooling.* The dissolved gases were expelled from the liquid , and accumulated in the upper part of the apparatus .
They could be collected by increasing the heat under the flask , so as to wash them out with steam into a test tube inverted over the recurved end of the narrow delivery tube .
It was found that the nitric acid solutions invariably gave some nitric oxide , and a much smaller quantity of nitrogen .
I have not paid special attention to the origin of this nitrogen , but it is certainly not atmospheric , as the absence of neon in the helium collected proves .
It may be due to * A fan was sometimes used to cool the tube , but without much advantage .
decomposition of nitric acid by radioactive bodies , but I have not tried whether it can be got from nitric acid only .
The gas collecting tube was half filled with oxygen , and some caustic potash floated on the surface of the mercury .
Thus the nitric oxide was oxidised and absorbed .
The excess of oxygen , carrying nitrogen and helium , was drawn into a gas burette and thence passed into a quartz tube full of mercury with melted phosphorus at the top .
The oxygen was burnt , and nitrogen , containing a small proportion of helium , remained .
The flasks usually contained about 3 litres of solution , in which about 600 grammes of thorianite was dissolved .
The nitrogen from each of them amounted to a fraction of a cubic centimetre , and served conveniently to carry the helium , in itself too little for pneumatic trough manipulation .
The final isolation of helium was effected by cooled charcoal .
The apparatus used is shown in fig. 2 .
The gas is introduced into a gas pipette a with a twoHon .
R J. Strutt .
Rate at which Helium [ July 23 , way stopcock , and thence into the exhausted apparatus .
It is followed up with mercury as far as the point b. After standing for some time over the cooled charcoal in c , it is drawn into the large vessel d , where it comes into contact with more cooled charcoal in the annexe ; is then heated to drive off any occluded gas , and allowed to cool .
Ultimately/ is closed , and d is fdled with TO MERCURY PUMP mercury so as to compress the residual helium into the capillary .
The arrangements for doing this are sufficiently obvious from the figure ; is a piece of very fine thermometer tubing 0T64 mm. diameter .
The upper end , instead of being fused up , was closed by drawing a thread of melted sealing-wax into it .
This gave an approximately plane end to the tube at right angles to its length , so that the volume of gas was calculable from the measured length it occupied , and the pressure .
The latter was measured by taking the 1910 .
] Produced in Thorianiteand Pitchblende , etc. 385 difference of level between the mercury in h and in by a cathetometer , and subtracting it from the barometric height.* The capillary depression in g ( 4'9 cm .
) was allowed for .
The actual pressure was about 20 cm .
of mercury , and the length of tube occupied by the largest volume of helium generated under observation 7'3 mm. The length by the column of gas was measured by a microscope with ocular scale , or , when too long for that , by a mirror glass scale , estimating to 1/ 10 mm. The apparatus lent itself very conveniently to spectroscopic observation of the gas , as a check on purity .
Two strips of tinfoil on the capillary , at 2 or 3 cm .
apart , were connected to an induction coil , and the pressure was reduced by lowering the mercury until the gas filled that part between the strips .
When the helium was pure a bright yellow glow was observed , which became bluish with the slightest admixture of nitrogen .
The spectroscope was , indeed , scarcely needed .
It was observed that if any trace of nitrogen remained it was rapidly removed by running the discharge for a short time , notwithstanding the absence of metallic electrodes , to which such actions are usually attributed .
The volume of gas was , of course , diminished by this process ; but , when once the nitrogen was gone , no amount of further sparking would diminish the volume of helium .
A correction to the measured volume is necessary , on account of the small quantity of helium which necessarily remains in the two charcoal vessels .
This fraction was determined , once for all , by starting with pure helium introduced into cl and measured .
Connection was then made to c and c , both previously exhausted and cooled .
Cutting off these connections , the volume was measured again in g. It was found to be diminished by about 1/ 10 part .
The validity of the measurements depends , of course , on the assumption that no part of the very small volume of helium obtained is absorbed by the charcoal , under the conditions of the experiment .
Although the absorption of helium at low pressures has been shown by Dewar and others to be very small , it was thought advisable to make direct tests .
This was done as follows : A small quantity of helium was very largely diluted with oxygen , so as to form a standard mixture ( of arbitrary composition ) .
This was placed in a graduated tube , and 1 c.c. of it drawn into the apparatus .
The helium contained measured 5'5G x 10-6 c.c. , a volume comparable with the smallest obtained in the helium experiments .
Seven cubic centimetres more of the mixture was then taken and most of the oxygen removed with phosphorus .
The residue was added to the gas already in the apparatus , and the total helium ( from 8 c.c. of mixture ) measured as 4-73 x 10~5 c.c. , 8-5 times that * h was open to the atmosphere during the volume measurements .
386 Hon. R. J. Strutt .
Rate at which Helium is [ July 23 , from 1 c.c. This proved that no appreciable absorption took place .
For , if there had been a slight absorption , it would have told much more on the small volume , and would have disturbed the ratio .
Further evidence on this point will be given immediately .
S 3.\#151 ; Results of First Seines.\#151 ; Thorianite from Galle ; the variety rich in uranium .
Three flasks ( numbered 4 , 5 , 6 ) set up , each containing 680 grammes in solution .
Experiment .
Volume of helium .
Initial blank test of all three flasks , P3 invisible After 205 days ' standing\#151 ; Flask No. 4 C.C. X 10-6 .
\lt ; 1-0 15 -5 34 -9 45 -9 2-9 24 -3 \lt ; 1-0 * Flask No. 4 +No .
5 Flask No. 4 + No. 5 + No. 6 Blank test of all three flasks , Ds conspicuous After standing 129 days more , all three flasks Final blank test , D3 invisible From these experiments\#151 ; f 8*18 x 10~5 c c Rate of production from all three flasks , per annum \#166 ; \lt ; g.gg ^q_5 c'c ' Mean ... ... ... 7'54 x 10~5 c.c. Rate of production per gramme of Galle thorianite , per annum ... ... ... ... ... ... ... ... ... ... ... ... ... 3-7 x 10~8 c.c. Helium initially present , per gramme ... ... ... ... . .
9-3 c.c. Time required to produce this ... ... ... ... ... ... ... 2'50 x 108 years .
Second Series.\#151 ; Ordinary thorianite containing 13T0 per cent. U3O8 , and 72'65 per cent. Th02 .
Three flasks ( numbered 2 , 3 ) set up , each containing 510 grammes .
Experiment .
Volume of helium .
Initial blank test of all three flasks , D3 doubtful c.c. x 10~6 .
\lt ; 0-5 After standing 141 days*\#151 ; Flask No. 1. .
5-0 Flask No. 1 + No. 2 11 3 Flask No. 1 + No. 2 + No. 3 16-5 Final blank test of all three flasks \lt ; 0-5 * As a matter of fact , the flasks were not all boiled out at quite the same time on this occasion .
The numbers given are slightly corrected to compensate for this , so that the results can be studied more easily .
1910 .
] Produced in Thorianiteand , etc. 387 From these experiments\#151 ; Rate of production from all three flasks per annum 4'27 x 10-5 c.c. Rate of production per gramme of ordinary thorianite 2*79 x 10~8 c.c. Helium initially present , per gramme ... ... ... ... .
7*8 c.c. Time required to produce this ... ... ... ... ... ... . .
2-8 x 108 years .
Pitchblende , from Joachimsthal , 353 grammes .
Dissolved ( see above , p. 382 ) in two separate flasks , which were always boiled at the same time , and the gases treated together .
Experiment .
Volume of lielium .
Blank test , Dq invisible c.c. X 10~6 .
\lt ; 0-5 2 -0 9-0 After 61 days ( not a good experiment ) After 294 days ( good experiment ) The last experiment is the only one on which stress can be laid .
It gives Helium per gramme pitchblende , per annum ... ... ... . .
3T6 x 10 ~8 .
The pitchblende experiments were not carried so far as those on thorianite , on account of the much greater difficulties of preparing the solution .
The best experiment , however , was a very satisfactory one .
The helium measured may be exhibited graphically as a function of the Number of flasks .
Fig. 3.\#151 ; ( 1 ) Gall thorianite .
( 2 ) Ordinary thorianite .
388 Bate at which Helium is Produced etc. number of flasks emptied , in those cases where the contribution of each was added successively , and shows fair proportionality ( fig. 3 ) .
This is additional evidence that no appreciable helium was lost in the charcoal , otherwise the first flask would have apparently yielded less than its due contribution .
S 4.\#151 ; Discussion of Results .
It now remains to compare the rate of formation of helium observed with that calculated theoretically .
For the calculation I refer to ' Boy .
' Soc. Proc. , ' A , vol. 81 , p. 276 , and vol. 83 , p. 97 .
Mineral . .
U8Og .
Per cent. Th02 .
Per cent. Helium production per gramme per annum .
C.c. x 10~8 .
Observed .
Calculated .
Thorianite .
Grille district ... 24 -50 65 -44 3-70 3-46 Ordinary thorianite 13 -10 72 -65 2-79 2-55 Pitchblende 37 -6 None 3-16 3 -44 The calculation is clearly justified by the direct observations , and can in future be employed with confidence in its substantial correctness .
The present experiments leave , I think , no doubt whatever that some specimens of thorianite are as much as 280 million years old .
I will take this opportunity of summarising the data in my previous papers* with regard to the duration of geological time , as deduced by the indirect method .
With the verification of this method now presented , I feel justified in stating them numerically without the qualifications before insisted on .
Mineral .
Geological horizon .
Minimum age .
Sphaerosidenite from Rhine provinces Haematite , Co. Antrim Oligocene Eocene Years .
8,400,000 31,000,000 150.000 .
000 710.000 .
000 Haematite , Forest of Dean Sphene , Renfrew Co. , Ontario Carboniferous limestone ... Archaean These are minimum values , because helium leaks out from the mineral , to what extent it is impossible to say .
Summary.\#151 ; The rate at which helium has been and is being produced in thorianite has been measured directly with reasonable accuracy , and is found to be in agreement with the rate calculated indirectly .
The helium now found in one sample examined would take 280 million years to accumulate .
Similar measurements have been made with pitchblende .
* ' Hoy .
Soc. Proc. , ' A , vol. 81 , p. 272 ; vol. 83 , p. 96 ; vol. 83 , p. 298 .

