On the contrary, in our experiments a beam which is propagated in the air contains at the greatest distances accessible to observation about 9/10 of β-rays, and the same is the case when the source of radiation is enclosed in a little sealed glass vessel. In M. Villard’s experiments, these deflected and penetrating β-rays did not affect the photographic plates beyond the first, because they are to a great extent diffused in all directions by the first solid obstacle encountered, and no longer form a beam. In our experiments the rays given off by radium and transmitted through the glass of the vessel were also probably scattered by the glass, but the vessel being very small would itself act as a source of β-rays at its surface, and we were able to follow the course of the latter to a great distance from the vessel. The cathode rays of Crookes tubes can only traverse very thin screens (aluminium screens of 0·01 m.m. thickness). A beam of rays striking the screen normally is scattered in all directions; but the diffusion becomes less with diminishing thickness of the screen, and for very thin screens the emerging beam is practically the prolongation of the incident beam. The deflected β-rays of radium behave in a similar manner, but the transmitted beam experiences, for the same thickness of screen, a much slighter modification. According to the experiments of M. Becquerel, the very readily deflected β-rays of radium (those with a relatively small velocity) are powerfully scattered by an aluminium screen of thickness 0·1 m.m.; but the penetrating and less deflected rays (rays of the cathode kind of great velocity) pass through this screen without being sensibly diffused, whatever be the inclination of the screen to the direction of the beam. The β-rays of great velocity penetrate without diffusion a much greater thickness of paraffin (several centimetres), and in this the curvature of the beam produced by the magnetic field can be traced. The thicker the screen, and the more absorbent the material of which it is composed, the greater is the modification of the deflected primitive beam, because, with increasing thickness of screen, diffusion occurs progressively among fresh groups of rays of increasing penetration. The β-rays of radium experience a diffusion in passing through the air, which is very marked for readily deflected rays, but which is much slighter than that produced by equal thicknesses of solid substances. For this reason, the β-rays traverse long distances in the air. _Penetrating Power of the Radiation of Radio-active Bodies._