190 
5. Dr. Joule when he demonstrates that there is no 
variation in economy, whatever the arrangement of the 
conducting metal, or whatever the size of the battery, 
demonstrates too much ; for it would follow from this, that 
no diminution in the size of the battery would affect the 
result injuriously. Why then should he prescribe a battery 
of such a size that the resistance may be neglected ? Take, 
again, his own figures in the 4th vol. of “ Sturgeon’s Annals.” 
By reducing his battery from 40 pairs to 10 he increased 
his economic duty from 65 00 to 97'00 or 50 per cent. 
6. But in reality Dr. Joule’s views in 1840 were more 
correct than they are now; for he was then not so far 
committed to the theory of the mechanical equivalence of 
heat. In 1840 he allowed that an economy could be effected 
by increasing the quantity of wire. Let us see whether he 
was right then, or whether he is so now. In his experi- 
ments in 1845 he used twice as much wire in his first two 
experiments as in his four last. What was the consequence ? 
In his first two experiments the average of foot-pounds per 
hour was 19,460, and duty per grain of zinc 98 - 35. In the 
four last experiments the average per hour 10,125, and duty 
per grain of zinc 51 '8, thus clearly showing that doubling 
the quantity of wire nearly doubled both the actual work 
done, and the economical duty. 
7. Again, a comparison of his two first experiments would 
tend to prove that reducing the intensity below one half 
reduces not only the work done per hour, but the duty 
also, instead of increasing it. For the full current being 
2,232, a reduction to 920 produced 21,100 foot-pounds of 
work per hour, and a duty of 1029 ; whereas a reduction to 
850 (by increasing the turns from 140 to 180 per minute) 
produced only 17,820 foot-pounds of work per hour, and a 
duty of only 93’8, or nearly 10 per cent. less. 
8. Dr. Joule’s accurate and most valuable experiments 
are still the great storehouse of facts on this subject. What 
I contend is, that they disprove, not prove, his own theory. 
