458 APPENDIX 



with KCl, thus effectively preventing any mixing of the two solu- 

 tions and diminishing any contact-difference of potential between 

 them.* The gas is passed through the electrode which dips into 

 the solution containing protein (the solution, that is, of unknown 

 hydrogen- or hydroxyl-concentration) at the rate of from one to 

 two large bubbles per second, and the excess of gas is allowed to 

 pass through the other electrode. The whole chain is immersed 

 in a small water-bath which is placed inside an incubator main- 

 tained at a temperature lying between 31 and 32 degrees (Cf. 

 below). It was thought necessary, at first, not to permit the 

 hydrogen to escape into the incubator, lest it should be ignited, 

 on opening the door of the incubator, by the flame beneath ; con- 

 sequently the electrodes were inserted into the half-elements 

 through tightly-fitting rubber stoppers, and rubber tubes were 

 attached to the side-tubes of the half-element and carried outside 

 the incubator and the cupboard within which the incubator is set 

 up. For reasons which will shortly be described, however, this 

 procedure was, of necessity, abandoned and the gas was permitted 

 to escape into the incubator. There appears to be no danger 

 involved in this procedure. The incubator is of the usual double- 

 walled type employed by bacteriologists; its internal dimensions 

 are 45 cm. wide by 24 cm. deep by 48 cm. high. It is provided 

 with two doors, the outer of the usual double-walled type, the 

 inner a glass door through which thermometers, etc., can be read 

 without disturbing the apparatus or causing fluctuations of tem- 

 perature by currents of air. The inner chamber is provided at 

 the top with two small air-exits. 



The potentials between the electrodes of the chain are measured 

 onalOO-cm. potentiometer bridge-wire (previously standardized).! 

 For the detection of the zero-point on the bridge-wire I have 

 employed a D'Arsonval galvanometer provided with a damping 

 coil; this instrument gave a decided deflection with the potential 

 corresponding to 1 mm. displacement on the bridge in ah of the 

 experiments described in Chap. IX. The constant fall in potential 

 from end to end of the potentiometer wire is best supplied by a 

 good storage-cell; but I have found an arrangement of four 

 Gladstone-Lalande cells (Model G-50), two in parallel and two 



* N. Bjerrum, Zeit. f. physik. Chem., 53 (1905), p. 428. 

 t The potentiometer manufactured by the Leeds & Northrup Co. is, how- 

 ever, more convenient to use, and also more accurate. 



