It has been shown, also, that a recently prepared salt of radium sets free a relatively small amount of heat. The heat set free in a given time then increases continually, and tends towards a definite value which is not quite reached at the end of a month.

When a salt of radium is dissolved in water and the solution enclosed in a sealed tube, the amount of heat given off at first is slight. It then increases and tends to become constant at the end of a month. When the limit is reached, the amount of heat given off from the radium in solution is the same as if it were in the solid state.

The amount of heat set free by radium at different temperatures can also be estimated by causing it to boil a liquefied gas and measuring the volume of gas that is liberated. The experiment can be performed with methyl chloride, which boils at -21°.

Figure I

The experiment was also performed by Dewar and myself with liquid oxygen (-180°) and liquid hydrogen (-252°). The latter is especially suitable for the experiment. A tube,

, closed at the lower end and surrounded by a Dewar vacuum-bulb, contains a little liquid hydrogen,

. A delivery-tube,