402 
Proceedings of Royal Society of Edinburgh. [july is. 
Similarly, the values in Table XI. are (as is shown by the fourth 
column) in fair agreement with the formula 
-5 = (182 - #)(8'58 + -03150), 
Kjt 
D 
so that - — vanishes at 182° C. (and would also vanish at - 208 *8° C.). 
Q^t 
The directly observed neutral point (by heating up junction) was 
190°-0 C. 
Let us consider now the interpretation of the first clearly proved 
result, viz., that -5_ (corrected) has the same value for a given 
C t 
temperature whether the battery current goes through the FeZn 
pile or the FeArg pile. (Let it be noted that all the corrections 
and all the conditions were the same in the corresponding X and Y 
measurements, so that this result is definitely proved.) 
Let the absolute temperature = t; 
the Peltier effect of FeZn = tf(t) ; 
the Peltier effect of FeAr g = t<f>(t) ; 
the thermoelectric power of FeZn = F(£) ; 
and that of FeArg, which by measurement is known 
to be constant, = a. 
The above experimental result becomes 
«/(0 = ■#>(«)• F ( / ) • 
■ • (i). 
And this is found to be still true (at least to within about 1° C.) 
when <h(t) = 0, i.e ., the Peltier effect, tf(t), vanishes at the neutral 
point.* 
The interpretation of equation (1) is that the Peltier effect in 
FeZn is equal to the product of the absolute temperature , the 
thermoelectric power, and some function of the absolute temperature 
which is the same for all pairs of metals, f That is — 
m 
m 
If it be taken as proved (and it has been to a certain extent) that 
* 183° C. (and not 190° C.) is clearly the true neutral point of the FeZn 
piles, since at 183° their thermoelectric power vanishes. 
t Similar measurements with two thermopiles of FeArg and NiArg also 
warrant this conclusion. 
