repetition of M, Arago's experiments on magnetism, ^c, 479 
the compass was observed, the former metal was as invaria- 
bly found to be placed below the latter, the other metals 
retaining their order. A possible explanation of this anomaly 
(should future experiments show that the fact depends on no 
fallacy) may be found in the principles hereafter to be 
explained, but we wish to be understood as speaking with 
reserve on this point. 
17. The following table is constructed in the same way 
as Tables I. II. III. with the addition only of the accelerating 
forces deduced on the supposition of uniform acceleration 
from the expression 
TABLE VI. 
N° of revo- 
lutions or 
parts, 
s — 
T] 
Copper 
mes of th( 
Zinc 
fir perforr 
Tin 
nance. F 
Lead 
or 
Antimo. 
t — 
Forces 
Copper 
f- 
deduce 
Zinc 
/ = 
d from 
10000c 
Tin 
the exp 
s 
)0 — 
Lead 
/-- 
ression 
Antim. 
f = 
0.25 
0.50 
0. 75 
1 . 0 
38.3 
54.2 
68.5 
79.8 
36.1 
SI -7 
63-9 
74.0 
74-8 
92.8 
107.8 
70-9 
102.5 
128.0 
151.2 
109.6 
157-9 
197-4 
232.4 
170 
1 70 
160 
157 
192 
187 
184 
183 
93 
89 
87 
86 
42 
48 
46 
44 
21 
20 
19 
19 
2. 0 
3. 0. 
4. 0 
5. 0 
1 10.6 
136.9 
1 60.0 
180.4 
106.2 
131-4 
152.8 
172.8 
156.8 
195-S 
229.5 
260,3 
221.8 
281.3 
335-0 
385.6 
351-7 
460.7 
164 
160 
156 
153 
177 
174 
171 
167 
81 
78 
76 
74 
41 
38 
36 
34 
16 
H 
Mean of all - - 
Mean of first six - - 
161 
179 
83 
41 
— 
163 
18 
The effect of torsion, resistance and friction, is very evident 
in the apparent diminution of the accelerating force in each 
revolution, so that only the numbers in the same horizontal 
lines can be regarded as comparable. Comparing accordingly 
the means of all for copper, 'zinc, tin, lead, and of the six first 
