1891-92.] Prof. Knott and A. Shand on Magnetic Strains. 251 
In all the tubes, and especially in Tube V. of narrowest bore, 
another interesting illustration of after-effect was observed. After 
a given field had been applied and removed several times in succes- 
sion, the meniscus in the capillary tube, which by its motion 
measured the change of volume, first moved slightly backward, and 
then turned and completed its total forward motion. On the first 
application of the same field in the reverse direction this backward 
“flash” w^as not observed. But if we continued to remove and 
apply the field in its new direction, this initial spring-back unmis- 
takably declared itself at each application of field. The pheno- 
menon is, in fact, essentially the same as that already described, and 
depends upon the current and its associated field not attaining their 
maximum values at once. With the powerful electro-magnet, which 
of necessity formed part of the circuit, the current, and therefore 
the field, rose in value sufficiently slowly to permit of the small 
initial values having their peculiar effect upon the biassed iron. 
In the final series of experiments on all the tubes, magnetic bias 
was removed as far as possible, and every field was applied in both 
directions. The very close agreement between the dilatations produced 
by a given field, taken first in one and then in the other direction, 
was a sufficient demonstration of the absence of any strong bias. 
The broad results are embodied in the following table, which 
gives the dilatations in all the tubes for certain definite fields. 
Any one column, excepting the first, relates to one tube. Each 
row gives the dilatations due to one particular field, whose value is 
entered to the left in the first column : — 
Table of Cubical Dilatations x lOh 
Tube. 
FieldT>\^ 
I. 
II. 
III. 
IV. 
V. 
50 
-3*6 
-1-5 
- 2 
- 1 
4- 1-3 
100 
-3 
-6-5 
- 9-5 
- 4 
4- 0-2 
150 
-f 0-4 
-7-2 
-18-6 
-12 
- 8-0 
200 
4-2 
-5-7 
-17 
-20 
-12 
300 
-1-3 
- 4-4 
-12-6 
-24 
-19 
500 
4-3-5 
-3-7 
-14 
-31 
-34 
700 
4-3-2 
-3 
-15-8 
-40-5 
-51-5 
1000 
4-3-8 
-3-3 
-17-8 
-46 
-84 
1400 
4-3-8 
-3 
-20 
-53 
-129 
The full significance of these numbers cannot be judged of till we 
