The cause of this effect was, however, rendered very evident by a few well-considered experiments made by Godlewski. The uranium nitrate was dissolved in hot water in a flat dish, and allowed to crystallize under the electroscope. Up to the moment of crystallization the β ray activity remained constant, but as soon as the crystals commenced to form at the bottom of the solution the β ray activity rapidly rose in the course of a few minutes to five times the initial value. After reaching a maximum, the activity very gradually decreased again to the normal value. If, however, the plate of crystals was reversed, the β ray activity was found at first to be much smaller than the normal, but increased as fast as that of the other side diminished. The explanation of this effect is simple. Ur X is very soluble in water and, at first, does not crystallize with the uranium, but remains in the solution, and, consequently, when the crystallization commences at the bottom of the vessel the upper layer of liquid becomes richer in uranium X. Since the β rays arise only from the product Ur X and not from the uranium itself, and the Ur X is mostly confined to the upper layer, a much greater proportion of the β rays escape than if the Ur X were uniformly distributed throughout the thick layer of uranium. When the amount of water added is just sufficient to supply the water of crystallization, the Ur X in the upper layer of crystals gradually diffuses back through the mass and, in consequence, the activity of the upper surface diminishes and of the lower surface rises. A similar explanation applies to the effects observed by Meyer and Schweidler. The water fraction, left behind after treatment with ether, contained all the Ur X. The first layer of crystals formed in it contained some Ur X, and this was for the most part confined to the top layer of crystals. The amount of β rays at first diminished owing to the gradual diffusion of the Ur X from the surface. In the first experiment, the amount of Ur X present was in radio-active equilibrium with the uranium, and, after the initial drop, the β ray activity remained constant. In the second experiment, the gradual rise is due to the fact that the crystals of uranium first formed contained less than the equilibrium amount of Ur X. After falling to a minimum, the β ray activity, in consequence, slowly rose again to the equilibrium value. These effects exhibited by uranium are of great interest, and illustrate in a striking manner the difference in properties of Ur X and the uranium. The gradual diffusion of the Ur X throughout the mass of crystals is noteworthy. By measurements of the variation with time of the β ray activity, it should be possible to deduce its rate of diffusion into the crystallized mass. Transformation products of Thorium.