298 Prof. W. J. Sollas. 



Copper. At. w., 63'8 ; sp. gr., 8'945 ; at. v., 7'0766. Volume of 



atomic sphere, 3'7053; diameter, 1'92. 

 Gold. At. w., 197-2; sp. gr., 19'33 at 17'5; at. v., 10-202. 



Diameter of atomic sphere, 2'169. 

 Iron. At. w., 56 ; sp. gr., 7'85 at 16 (Caron) ; at. v., 7134. 



Diameter of atomic sphere, 1*925; or sp. gr. 8'139 (Roberts 



Austen) ; at. v., 6'88 ; diameter of atomic sphere, T902. 

 Manganese. At. w., 55; sp. gr., 7'3921 at 22; at. v., 7'44. 



Diameter of atomic sphere, 1*952. 

 Platinum. At. w., 194-8; sp. gr., 21'5 at 17'6 ; at. v., 9'0605. 



Diameter of atomic sphere, 2'085. 

 Palladium. At. w., 106; sp. gr., 11'4 at 22'5; at. v., 9'2983. 



Diameter of atomic sphere, 2' 103. 



The Absorption of Hydrogen by Palladium. Strong confirmatory 

 evidence of the existence of the open packing which we have assigned 

 to the metals crystallising in the cubic system is afforded by the 

 phenomenon of solid solution (so called), and particularly by the 

 absorption of hydrogen by palladium. "When similar spheres are 

 arranged in open cubic order, they form straight rows in contact, 

 running parallel to the edges of the cube they constitute, and corre- 

 sponding to these files of spheres are open galleries, lying between 

 and running parallel with them. Through these galleries atoms, if 

 small enough, might pass from end to end without encountering any 

 obstacle, and thus the transpiration of hydrogen through metallic 

 plates might be explained. Further, between every set of eight 

 atoms, forming a primitive cube of the pile, the gallery widens out 

 into a chamber, in which an atom smaller than that of the metal 

 might conceivably lodge. The diameter of an atom which could 

 occupy the space between eight atoms, forming a primitive cube, of 

 palladium, can readily be calculated. The diameter of an atom of 

 palladium has already been determined to be 2' 103, the edge of a 

 cubelet formed of eight atoms is therefore 4'206, and the length of 

 the trigonal axis of such a cube is 4'206 x \/3 = 7'285 ; and (7'285 

 4'206)/2 = 1-538, the length of the diameter of an atom, which 

 would just occupy the central space. This estimate, however, 

 requires modification, by virtue of the fact that palladium progres- 

 sively increases in volume as the absorption of hydrogen takes place. 

 In Dewar's determinations the expansion was measured by the 

 change produced in the specific gravity of the palladium; the lowest 

 specific gravity which Dewar observed was 10'8033 ; this gives for 

 the edge of the primitive cube a value of 4'2818. Assuming that 

 the atoms of palladium have not increased in volume by absorbing 

 energy, but have simply become more remote from one another, we 

 may proceed as follows : 4'2818 X -/3 =. 7'416, the length of the tri- 



