gen wend their way would never be sufficiently high to enable them to quit the earth altogether, and consequently we might in such a case expect to find our atmosphere largely charged with hydrogen. Considering the vast abundance of hydrogen in the universe, it seems highly probable that its absence from our air is simply due to the circumstances we have mentioned. In the case of a globe so mighty as the sun, the attraction which it exercises, even at the uppermost layers of its atmosphere, is so intense that the molecules of hydrogen never attain pace enough to enable them to escape. Their velocity would have to be much greater than it ever can be if they could dart away from the sun as they have done from the earth. It is not, therefore, surprising to find hydro* gen in the solar atmosphere. In a similar manner we can explain the abundance of hydrogen with which the atmospheres of other massive suns like Sirius or Vega seem to be charged. The attraction of these vast globes is sufficiently potent to retain even an atmosphere solely composed of this subtle element.
It is now easy to account for the absence of atmosphere from the moon. We may feel confident from the line of reasoning here followed that neither of the gases, oxygen or nitrogen, to say nothing of hydrogen, could possibly exist in the free state on a globe of the mass and dimensions of our satellite. The pace with which the molecules of oxygen and nitrogen speed on their way would be quite sufficient to render their abode unstable if it should ever have appeared that circumstances placed such gases on the moon. We need, therefore, feel no surprise at the absence of any atmosphere from the neighbouring globe. The explanation is given by the laws of dyna-
