396 
Proceedings of Royal Society of Edinburgh. [july 18, 
the opposite direction for 100, and broken again for 100 more. 
These periods were chosen so that the deflection might become 
nearly constant before the current was broken. The current, which 
was from 2 or 3 Bunsen cells, was very nearly constant throughout 
the 200 seconds. The curves (1), (2), and (3) in the diagram are 
drawn from these tables, the abscissa being the time (in half-seconds) 
and the ordinate the galvanometer deflection. 
Table VI. Curve (1). 
Time, 
A Seconds. 
Deflection. 
Time, 
A Seconds. 
Deflection. 
Time, 
\ Seconds. 
Deflection. 
Time, 
s Seconds. 
[Deflection. 
0 
0 
92 
1660 
1 
Reverse 
283 
-1590 
3 
100 
94 
1680 
200 ) 
current 
287 
-1600 
4 
200 
99 
1690 
( 
made. 
293 
-1610 
5-5 
300 
100 | 
Current 
201 
100 
298 -5 
-1620 
7 
400 
broken. 
204 
0 
300 | 
Current 
8 
500 
104 
1600 
205 
- 100 
broken. 
10 
600 
106 
1400 
207 
- 200 
302 
-1600 
11 
700 
111-5 
1000 
210 
- 400 
304 
-1500 
13 
800 
113 
900 
210 
- 500 
307 
-1300 
15 
900 
116 
800 
213 
- 600 
309'5 
- 1100 
18 
1000 
117 
700 
215 
- 700 
311 
-1000 
21 
1100 
121 
600 
217 
- 800 
313 
- 900 
25 
1200 
123 
550 
223 
-1000 
315 
- 800 
29 
1300 
125 
500 
227 
-1100 
317 
- 700 
33 
1350 
128 
450 
229 
-1200 
319 
- 600 
36 
1400 
130 
400 
234 
-1300 
323 
- 500 
40 
1450 
133 
350 
239 
- 1350 
325 
- 450 
45 
1500 
137 
300 
244 
-1400 
327 
- 400 
52 
1550 
143 
250 
246 
-1450 
329 
- 350 
53 
1560 
149 
200 
248 
-1470 
336 
- 300 
59 
1580 
162 
150 
257 
-1510 
340 
- 250 
61 
1590 
165 
140 
261 
-1520 
347 
- 200 
65 
1600 
167 
130 
263 
-1530 
355 
- 150 
67 
1620 
174 
120 
265 
- 1540 
376 
- 100 
72 
1630 
179 
110 
268 
-1550 
381 
- 50 
75 
1640 
187 
100 
274 
-1570 
390 
- 40 
78 
1650 
278 
-1580 
§ 3. Thermoelectric Power of Iron German Silver. 
As in all the above experiments the measurements were made by 
means of iron-German-silver thermopiles, it was highly important 
that the thermoelectric power of the iron and German silver used 
should be carefully measured throughout the range of tempera- 
tures at which the piles were used. Bor this purpose, the FeArg 
junction, instead of being heated in oil, was packed in asbestos 
