208 Proceedings of the Koyal Society of Edinburgh. [Sess. 
This table is composed of six large groups of four rows and eight 
columns, any one group giving the measurements associated with a 
particular value of the longitudinal field. The first row of each group 
gives the values of H, the second the values of T(H), the third the values 
of T, and the fourth the values of H(T). Thus, with the exception of 
the third row of each group, every row has to do with a particular 
value of the longitudinal field h. Similarly, with the exception of the 
numbers of the first row in each group, the numbers in any one column 
have to do with a particular value of the transverse field t. The values 
of the longitudinal field are entered down the left-hand margin, and 
the values of the transverse along the top of the table. The nature 
of the particular field or combination of fields is indicated in the second 
column. 
In all cases each field whose effect on the resistance was being studied 
was applied and removed twice in succession, with change of direction 
between the two applications. The nickel was thus subjected to a magnetic 
cyclic state as regards the magnetizing force which is being applied and 
removed. In the concomitant changes of resistance there was no evidence 
of hysteresis. Such effects, indeed, could hardly have been looked for in the 
case of the transverse field, seeing that the width of the nickel strip was 
short compared to its breadth as a cylindrically shaped coil. It was other- 
wise, however, with the field applied longitudinally along the coiled strip, 
which formed the core of the anchor-ring magnetizing coil. We cannot 
doubt that there was magnetic hysteresis of the usual kind ; but no 
hysteresis was observed in the cycle of resistance change accompanying the 
cycle of magnetization. 
A glance through the numbers tabulated in Table A shows that the 
resistance is increased in the direction of magnetization, and decreased in 
the perpendicular direction. This was established in 1856 by Thomson 
(Kelvin), and has been frequently verified by other experimenters. In the 
table, all the numbers in the H rows are positive, and those in the T rows 
are negative. The numbers in any one H row should be the same ; for 
they are the changes of resistance due to the action of the same field. 
Similarly, the T numbers in any one column represent the change of resist- 
ance due to the application of one particular transverse field, and should 
also be the same. It will be noticed that the numbers corresponding to 
each H or each T fluctuate somewhat, and this fluctuation may be taken as 
a guide to the possible errors of the experiment. 
In Table B the mean values of the different measurements of each 
H and of each T are recorded, with the average error attached. The 
