Purification of Gallium. 391 



longed undercooling which metallic gallium ordinarily shows. 

 Although we do not know the necessary conditions for the 

 generation of trees we have never failed to produce them at 

 will by cooling the electrolyte in an ice hath. If the solution 

 is at room temperature and a liquid globule is growing, the 

 change to the solid phase may he brought about gradually by 

 placing ice in the tray and slowly cooling the liquid, whereas 

 if the solution and electrodes are at 0° C. before and after the 



Fig. 2. 



current is started the deposit will be solid from the very begin- 

 ning. 



A fairly complete idea of the appearance of the trees may 

 he derived from the associated figures which were obtained by 

 first photographing the objects natural size with a long focus 

 Dallmeyer "Rapid Rectilinear" portrait lens stopped down to 

 " U. S. 100" and then enlarging the negatives nearly two-fold. 

 Figures 2a, 3a, and 3b show certain trees when the line of 

 sight was horizontal, while 2b, 3c, and 3d give the respective 

 aspects of the same trees when viewed from below, that is, 

 when looking vertically upward along the axes of the cathodes. 

 The three photographs first mentioned illustrate the fact that 

 the branches of the trees are concave on the upper side. 

 Figures 2a and 2b pertain to a tree which was formed with 

 only one cathode in the solution. The two longest branches 

 in 2b were directed approximately toward the outer edges of 

 the anode, as was often found to be the case for a single 

 cathode. Figure 3 represents two trees which were deposited 

 simultaneously on two cathodes joined in parallel. In 3a and 

 3b the trees appear as seen from the center of the submerged 

 portion of the anode, the plane containing the vertical axes of 

 the cathodes being normal to the line of sight. For 3c and 3d 



