1898-99.] Prof. Knott on Magnetic Tivist in Nickel Tubes. 593 
Table IV . — Comparison of Observed and, Calculated Tivists. 
Longitudinal 
Field. 
H' 
T 
= 2*4. 
Twist. 
H' 
T 
= 4-33. 
Twist. 
H' 
T 
= 6-3. 
Twist. 
H' = 
T 
15-8. 
Twist. 
20 
•78 
1-5 
1-41 
3-4 
2-08 
4’5 
4’6 
6-0 
40 
1-07 
2*8 
1-86 
47 
2-79 
6-3 
6’7 
10*7 
60 
1-15 
3-0 
2-07 
4-2 
3-00 
6-2 
7-48 
11-7 
80 
1-17 
2-2 
2-11 
4-0 
3-03 
5-0 
7-8 
11-0 
100 
1-08 
2-1 
2-01 
3-8 
2-96 
4-6 
7-38 
8T 
150 
•94 
1*5 
1-69 
3-2 
2-46 
3-5 
5-87 
7 T 
The curves showing the march of r with magnetic field for H' 
equal to 2*4, 6 - 3, and 15*8 are shown in dotted outline in fig. 1, 
and may be compared at a glance with the corresponding full-line 
curves of observed twist. 
There is an evident broad similarity. The two outstanding 
differences are : — 
(1.) The values of the observed twist for a given combination of 
fields are always greater than the values of the calculated twist, 
the difference being greatest for the lowest circularly magnetizing 
force and least for the greatest. 
(2.) The observed maximum twists occur in lower fields than 
those indicated in the calculated values, and are distinctly more 
pronounced. 
And to these may be added — 
(3.) The twist for a given combination of circular and longi- 
tudinal magnetizations is much less when the circularly magnetiz- 
ing force is superposed on the longitudinally magnetizing force 
than when the order is reversed. 
In seeking for an explanation of these quantitative discrepancies 
we must bear in mind two facts of the first importance. 
I. Because of seolotropy pre-existent or produced by the magne- 
tizing forces, the principal direction of strain does not probably 
coincide with the direction of the magnetizing force; 
