1892.] On the Electrolysis of Silver Nitrate in Vacuo. 345 



oxygen, for when the electrolysis is carried out in an atmosphere of 

 oxygen the deposits are smaller than those obtained in air. 



The apparatus we employed to obtain a deposit in vacua is illus- 

 trated in fig. 1. An inverted bell- jar, closed at the bottom by an 

 india-rubber stopper and at the top by a plate of glass, contains a 

 tight-fitting cylindrical cage of wire gauze, which serves as support 

 to the electrodes. The platinum basin is placed on two stout copper- 

 wires, F, which are soldered to the cage. Metallic contact between 

 the wires and the bowl is secured by the help of tinfoil, which is 

 wrapped round the wires and forms a cushion on which the bowl 

 rests. One of the wires leading to the battery is soldered to the 

 cage. The anode is suspended from a glass rod, C, fixed to the cage 

 near its upper end, the current being conveyed to the anode by an 

 insulated wire passing through a glass tube, B, which is also secured 

 to the cage. Three pieces of glass tubing pass through the india- 

 rubber stopper; one serves to exhaust the vessel, while the wires 

 leading to the battery pass through the remaining two. 



The stopper is rendered air-tight by means of Faraday cement, and 

 some grease has to be used to prevent leakage between the glass 

 plate and bell-jar. To prevent particles of this .grease contaminating 

 the solution, a tightly- fitting piece of cardboard, not shown in the 

 figure, was placed above the cage. In the latter part of the investi- 

 gation two nearly identical bell-jars were used. 



The same current always passed through two or three voltameters 

 in succession, and the deposits obtained simultaneously were com- 

 pared with each other. One of the platinum bowls, to be referred to 

 as the large bowl, has a diameter of 5 inches, while the smaller bowl 

 had a diameter of 3| inches. 



The silver anodes had a thickness of about 2 mm., and generally 

 larger anodes were used in the large bowl than in the small one. 

 With respect to the contact between the anodes and the platinum 

 wires conveying the current, it seems worth while to draw attention 

 to a precaution, which, if neglected, may cause serious trouble. We 

 placed at first, for the sake of convenience, the anodes simply into 

 two loops of platinum wire. These loops crossed at right angles as 

 in fig. 2. The current under these circumstances is apt to pass 

 partly from the platinum wires, and dark red crystals (probably 



FIG. 2. 



