666 PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 
Experiment IV. November 19tb, 1853. 
(Current seven times eight minutes i 
differences of temperature after 
By end next A. 
ti each direction.) Temperatures and 
;ight minutes of current entering 
By end next B. 
Augmentations of 
differences from 
middles to ends of 
periods. 
Periods. 
Ta- 
Tb- 
Tb-Ta = D- 
Ta- 
Tb- 
Tb-Ta- 
D'-D. 
I. 
5^50 
59-30 
i-80 
58-02 
59-84 
i-82 
°-02 
IL* 
48-20 
51-15 
2-95 
46-82 
49-79 
2-97 
-02 
III. 
46-49 
49-13 
2-64 
47-01 
49-95 
2-94 
•30 
IV. 
48-41 
51-69 
3-28 
48-31 
51-99 
3-68 
•40 
V. 
48-36 
51-74 
3-38 
48-18 
51-80 
3-62 
•24 
VI. 
48-00 
51-20 
3-20 
48-00 
51-49 
3-49 
•29 
Vll. 
48-31 
51-51 
3-20 
48-06 
51-60 
3-54 
•34 
Means for five periods.. 
49-3243 
52-2457 
3-14 
49-20000 
52-35143 
3-454 
•314 
Augmentation of difference during periods included... 0'57 
Deduct average augmentation per half-period *057 
Effect due to reversal of current 0‘'*2575 
in favour of Resinous Electricity. 
41. A full analysis of the differential variations throughout each of these experi- 
ments, derived from observations of the thermometer taken every quarter of a 
minute, was made in each case immediately after the conclusion of the experiment 
(see Tables I. and II. 47 below), and was sufficient to convince me that the true 
effect in the iron conductor is of the kind indicated by the preceding summary of 
the effects apparent at the ends of the periods. 
42. To try whether or not the very considerable effect thus discovered depended 
on some inequality in the conductor itself, I made an experiment on the 25th of 
November 1853 exactly like the two preceding, with the exception that the middle 
vessel previously used as a heater was filled with cold water at the commencement. 
The current was sent six times eight minutes in each direction ; the thermometers were 
noted every quarter of a minute; and the observations were reduced and compared 
in the usual way. The result gave no effect of the kind observed in the preceding 
experiments, but (probably because of a temporary failure in the water-supply for the 
coolers) showed, on the contrary, a deviation in the mean difference of temperature 
amounting to °‘029 Cent., being about a tenth part of the amount of that effect, but 
in the opposite way according to the direction of the current through the conductor. 
Before the experiment was concluded boiling water was poured through the central 
vessel and left filling it, but with no lamp below. The two thermometers (A and B) 
being thus raised to about 27° Cent., the current was again started and was sent 
through the conductor for three times four minutes in each direction. The thermo- 
meters rose each nearly 2°, but fell again by nearly 5^° before the conclusion. The 
mean differential result, whether from these three periods (amounting to °‘05 Cent.), 
* Rejected because of a failure in the water-supply through the coolers during the whole of Period I. 
