is of the right order of magnitude to be expected on the disintegration theory, if uranium is the parent of radium. The activity of pure radium, compared with uranium, is not known with sufficient accuracy to determine with accuracy the theoretical proportion of radium to uranium.
The production of radium from uranium, while very strongly supported by these experiments, cannot be considered definitely established until direct experimental evidence is obtained of the growth of radium in uranium. The rate of production of radium to be expected on the disintegration theory can readily be estimated. The fraction of uranium breaking up per year has been calculated (section 261) and shown to be about 10-9 per year. This number represents the weight of radium produced per year from 1 gram of uranium. The emanation, released from the amount of radium produced in one year from 1 gram of uranium, would cause an ordinary gold-leaf electroscope to be discharged in about half-an-hour. If a kilogram of uranium is used, the amount of radium produced in a single day should be easily detectable.
Experiments to detect the growth of radium in uranium have been made by several observers. Soddy24 examined the amount of emanation given off at different times from one kilogram of uranium nitrate in solution, which was originally freed from the small trace of radium present by a suitable chemical process. The solution was kept stored in a closed vessel, and the amount of emanation which collected in the solution was measured at regular intervals.
Preliminary experiments showed that the actual rate of production of radium was far less than the amount to be expected theoretically, and at first very little indication was obtained that radium was produced at all. After allowing the uranium to stand for eighteen months, Soddy states that the amount of emanation was distinctly greater than at first. The solution after this interval contained about 1·5 × 10-9 gram of radium. This gives the value of about 2 × 10-12 for the fraction of uranium changing per year, while the theoretical value is about 10-9.
Whetham25 also found that a quantity of uranium nitrate which had been set aside for a year showed an appreciable increase in the content of radium, and considers that the rate of production is faster than that found by Soddy. In his case, the uranium was not originally completely freed from radium.
Observations extending over years will be required before the question can be considered settled, for the accurate estimation of small quantities of radium by the amount of emanation is beset with difficulties. This is especially the case where observations are made over wide intervals of time.
The writer has made an examination to see if radium is produced from actinium or thorium. It was thought possible that actinium might prove to be an intermediate product between uranium and radium. The solutions, freed from radium, have been set aside for a year, but no certain increase in the content of radium has been observed.
There is little doubt that the production of radium by uranium first proceeds at only a small fraction of the rate to be expected from theory. This is not surprising when we consider that probably several changes intervene between the product Ur X and the radium. In the case of radium, for example, it has been shown that a number of slow changes follow the rapid changes ordinarily observed. On account of the feeble activity of uranium, it would not be easy to detect directly the occurrence of such changes. If, for example, one or more rayless products occurred between Ur X and radium, which were removed from