190 PROCEEDINGS OF THE AMERICAN ACADEMY 



The existence of a transverse electromotive force under the condi- 

 tions described might be accounted for in either of two general ways : 

 1st. The combined action of the main electric current and the magnetic 

 force might be supposed to produce in the conductor a state of strain, 

 giving it for the time being properties similar to those possessed by 

 certain crystals, which do not conduct ecpially well in all directions. 

 This might be called the static theory of the phenomenon, and one 

 adopting it would expect the transverse effect to increase or decrease 

 with the electric resistance of the conductor, that is, he would expect 

 to see a fall of temperature accompanied by a diminution of the trans- 

 verse effect, and vice versa. 2d. The transverse electromotive force 

 might be regarded as the result of molecular or cellular motions, prob- 

 ably rotations, set up within the conductor by the magnetic force, and 

 acting upon the main current of electricity in such a manner as to 

 produce a tendency to deflection. This mi<iht be called the kinetic 

 theory, and one holding it need not be surprised to find the transverse 

 electromotive force independent of those temperature change* which 

 affect electric resistance. 



As most observations upon the temperature changes of the Mall 

 effect had been confined to narrow limits, between 30° C. and 0° C, 

 it seemed important to undertake a serious and extended investiga- 

 tion reaching through a much greater range of temperatures. This 

 investigation is by no means completed, but some results of interest 

 have already been attained. Most of the experimeutal work leading 

 to these results was done by Mr. A. L. Clough, a graduate student, 

 in the Jefferson Physical Laboratory of Harvard College during the 

 academic year 1891-1)2. 



The substances used were, especially, soft cold-rolled steel and soft 

 rolled nickel, but some few observations were made with carbon, cop- 

 per, and phosphor-bronze. The general method of the experiments was 

 similar to that which Professor Hall had used in previous investigations. 

 The most important variations, which will presently be described, 

 were those required by the use of high temperatures. The substances 

 examined, with the exception of carbon, were in the usual shape of 

 thin strips several centimeters long and about one centimeter wide, 

 with narrow arms projecting at right angles with the sides. None of 

 the work lays claim to great numerical exactness. It is a pioneer 

 exploration rather than an accurate survey. 



In some observations at moderate temperatures the same narrow 

 water bath was used that was described in the American Journal of 

 Science, Vol. XXIX. p. 118. A hot-air bath was used for tempera- 



