426 Mr. E. H. Hall on the Rotation of the 
This table shows the results of experiments with the metal 
strips thus far described in detail. The last two columns give 
approximately the composition.of plates which would have the 
observed rotatory powers if copper and zinc in uniting retained 
their individual powers unaltered. It will be perceived that 
in every case the proportion of copper demanded upon this 
basis is greater than the proportion in the actual plate. 
At the time these experiments were made I supposed plate B 
to contain about 92 per cent. of copper, which would have 
furnished an exception to this rule. Thinking some serious 
error might have been made in estimating the thickness of 
this plate, or that of the copper plate, or the zinc, I prepared 
with very great care plates No. 2 of A, B, and G. These 
_ new plates were similar to those already described, but were 
in length between B and B about 22cm. As the vacation 
was now over, and my opportunities for experimental work 
were limited, I contented myself with making, in the main, 
comparative experiments with these plates. I did not re- 
determine the horizontal intensity of the earth’s magnetism, 
and I made no attempt to determine the direct effect of the 
electromagnet upon the Thomson galvanometer, with which 
the transverse current was observed. I knew that this direct 
effect was small; and I endeavoured to arrange matters in 
such a way that it would affect all the new results with A, 
B, and G in the same direction and to nearly the same extent. 
The intensity of the magnetic field in the experiments with 
these plates was about the same as in the previous experiments 
with the alloys. This intensity was determined, as usual, on 
October 20th, when the effect of change of temperature upon 
the plate B No. 2 was tested, but on October 13th, when 
A No. 2 and B No. 2 were compared, and on October 27th, 
when A No. 2 and G No. 2 were compared, this determina- 
tion was not considered necessary. On October 13th A No. 2 
was tested first, then B No. 2, then A No. 2 again. This 
arrangement tended to make the result of the comparison in- 
dependent of any progressive diminution in the strength of 
the current operating the electromagnet, and it saved much 
time. On October 27th the same method was followed. 
I shall not attempt to give the results of these experiments 
in absolute measure, but shall write the results obtained on 
an arbitrary scale. 
