The following table shows the results obtained for different currents of hydrogen and oxygen. Current of Gas _T₁_ _T₂_ Hydrogen ·25 c.c. per sec. −151·3 −150 „ ·32 „ „ −153·7 −151 „ ·92 „ „ −152 −151 „ 1·38 „ „ −154 −153 „ 2·3 „ „ −162·5 −162 Oxygen ·34 „ „ −152·5 −151·5 „ ·58 „ „ −155 −153 The temperature _T₁_ in the above table gives the temperature of initial volatilization, _T₂_ the temperature for which half of the condensed emanation had been released. For slow currents of hydrogen and oxygen, the values of _T₁_ and _T₂_ are in good agreement. For a stream of gas as rapid as 2·3 cubic centimetres per second the value of _T₁_ is much lower. Such a result is to be expected; for, in too rapid a stream, the gas is not cooled to the temperature of the spiral, and, in consequence, the inside surface of the spiral is above the mean temperature, and some of the emanation escapes at a temperature apparently much lower. In the case of oxygen, this effect appears for a gas stream of 0·58 cubic centimetres per second. In the experiments on the thorium emanation, on account of the rapid loss of activity, a slightly different method was necessary. The steady stream of gas was passed over the thorium compound, and the temperature was observed at the instant when an appreciable movement of the electrometer appeared. This gave the temperature at which a small fraction of the thorium emanation escaped condensation, and not the value _T₁_ observed for the radium emanation, which gave the temperature for which a small fraction of the previously condensed emanation was volatilized. The following table illustrates the results obtained. Current of Gas Temperature Hydrogen ·71 c.c. per −155° C. sec. „ 1·38 „ „ −159° C. Oxygen ·58 „ „ −155° C. On comparing these results with the values obtained for the radium emanation, it will be observed that with equal gas streams the temperatures are nearly the same. A closer examination of the thorium emanation showed, however, that this apparent agreement was only accidental, and that there was, in reality, a very marked difference in the effect of temperature on the two emanations. It was found experimentally that the radium emanation was condensed very near the temperature at which volatilization commenced, and that the points of condensation and volatilization were defined fairly sharply. [Illustration: Fig. 60.]