RESEARCHES IN CONTACT ELECTRICITY. 281 
EXPERIMENT IX. (April 4). 
(Fig. 4; symbol x ). 
Temperature. Deflection (Zn, | Zn,) 
65° C. —42 
45 —22°5 
42:7 —19°2 
40°6 —17°7 
38°8 —18°5 
28°6 —8 
In the diagram (fig. 4), two lines are drawn, each representing one of the 
above experiments. The dotted line is that which best agrees with the readings 
of experiment VIII., the points on the curve of which are represented on the 
diagram as “dots.” The curve-points of experiment IX. are entered as crosses, 
and they all lie very near the continuous line drawn on the diagram. The 
tangent of inclination of this line is—‘9, expressed in diagram units. 
Apparently, then, zinc varies more rapidly with temperature than iron ; and 
hence, since zinc is the more positive, the contact difference of potential between 
zinc and iron falls off, as both are simultaneously raised in temperature ; 
a result in accordance with the indications of the earlier experiments with 
zinc and iron when both were made to vary similarly in temperature. This 
suggested the possibility that the more positive metal might be subject to 
the greater temperature-variation. According to this hypothesis, tin, which 
_ occupies in the electromotive series a position intermediate to zinc and iron, 
should give a correspondingly intermediate line for its temperature-variation. 
It was impossible, however, with the means I had at my disposal, to arrive at 
anything like a quantitative result for tin. Not having at the time another tin 
surface, I was compelled to make use of either zinc or iron as the other con- 
denser plate ; and, as both of these gave large deflections with tin, the readings 
were wild and unsatisfactory. No experiment gave even self-consistent results; 
and no two of them had much in common—except the undoubted characteristic 
which indicated a similar “negative growth” with rise of temperature of the 
tin surface. 
As already noticed, the permanency of this negative-growth with temperature 
increase after the surface is cooled—a characteristic which was established by 
direct experiment in every case—proves conclusively that whatever change in 
the electromotive force of contact of any two of the metals, iron, zinc, copper, 
and tin, may be due directly to change of temperature ; such a possible change 
is quite inappreciable by ordinary contact methods, and is altogether masked 
by changes due to other and secondary causes. In seeking for such causes, we 
must consider the probable alteration with temperature in the density of the 
gaseous film condensed over the metal surface, which alteration, however, is not 
