126 PROCEEDINGS OF THE AMERICAN ACADEMY. 



between the moving and stationary parts. For this reason, the shafts 

 were made as small as possible, 3/16 inches in diameter, thus reducing 

 the relative velocity at the point of contact, mercury was chosen as 

 the contact substance, and the whole apparatus was made of the same 

 material, steel, throughout. Measurements of the electromotive forces 

 occurring in these mercury contacts, were made by driving the appa- 

 ratus with the contacts short circuited. The potential differences 

 found were in the neighjporhood of 1/20 millivolt. 



The Magneto. 



Returning again to Figure 5, T is a worm driving the shaft S 

 through a worm wheel not seen in the drawing. Measurements of the 

 speed of rotation of the apparatus were made by reading the voltage 

 produced by a magneto driven from this shaft S. This was a small 

 three-bar magneto manufactured a number of years ago by the Holtzer 

 Cabot Co. of Brooldine. This was the least satisfactory part of the 

 apparatus, since the voltage readings at a constant speed were liable to 

 small, sudden fluctuations largely due to poor contact at the commu- 

 tator. Both graphite and woven wire commutator brushes were tried. 

 The accuracy of the speed determinations, however, seemed to be of the 

 same order of magnitude as that of the other measurements, and it did 

 not seem advisable to procure one of the newer magnetos or any of the 

 more costly forms of apparatus for speed measurement. A small motor 

 was arranged for driving the magneto independently, and the magneto 

 was standardized after each series of measurements with the help of a 

 stop watch and suitable counter. 



The Electrical Measuring Instruments. 



The potential differences obtained were of the order of a few milli- 

 volts. For their measurements, a Leeds and Northrup pontentiometer 

 and a suitable galvanometer were used. A cadmium element supplied 

 by the Weston Electric Co. was- used as a standard cell, and this was 

 further compared with another Weston cell both at the beginning and 

 end of the measurements. In the actual experiments a reading of the 

 voltmeter which gave the speed of the apparatus, and the potentiom- 

 eter reading were taken as nearly simultaneously as possible. 



5. The Method of Procedure. 



With the apparatus described, measurements were made on solutions 

 of potassium, sodium, lithium, and hydrogen iodides. The solutions 

 contained exactly 1 mol of the salt and 1/100 mol of I2 in a kilogram 



