1898—99.] Prof. Knott on Magnetic Twist in Nickel Tubes . 587 
tube ; and merely a rough first approximation to concordance 
could be looked for when the formula was applied to a solid wire. 
It was a natural extension of my recent work on tubes to measure 
if possible the Wiedemann effect in them also. The comparatively 
large radius of even the narrowest tubes which had been found 
suitable for the other work precluded the hope of getting power- 
ful circular magnetizations, so that, according to the view advocated, 
the Wiedemann effect would be very small. As will be seen 
below, it was, in fact, little more than just measurable in a very 
thin- walled nickel tube. 
The nickel tube used in this investigation was that known as 
Cl in the paper just referred to. It was formed by coiling a piece 
of sheet nickel ‘027 cm. in thickness. The tube was 25 cm. long 
and 2-5 cm. external diameter. The elongations at the outer 
surface, when the tube is longitudinally magnetized, are given in 
Table V., p. 485, of my last paper cited above. For subsequent 
reference I reproduce certain of the results here. The quantities 
in the columns headed X, /x, v are the elongations multiplied by 
10 6 of a surface element; A. representing the elongation parallel to 
the axis of the tube, g the “tangential” elongation, and v the 
radial elongation. 
Table I. — Elongations in Various Magnetizing Fields of 
Nickel Tube Cl. 
Field. 
X 
[X 
V 
Magne- 
tization. 
25 
- 4 
+ 1-8 
+ 2*2 
202 
50 
- 9‘5 
+ 5-0 
+ 4*5 
315 
100 
-18*2 
+ 10*9 
+ 7*4 
392 
150 
-22-5 
+ 13*8 
+ 8*7 
430 
200 
-25 '4 
+ 15*6 
+ 9*8 
442 
300 
-28*2 
+ 17*1 
+ 11*1 
450 
(2.) Measurement of the Wiedemann Effect in the Tube. — The 
tube was set vertical in the heart of the magnetizing coil used in 
the former experiments in magnetic strains, the upper end being 
fixed* and the rest of the tube being left as free as possible* Hear 
