Changes in Temperature and Distribution of Magnetism. 281 
Tables of the results for magnets A and B are given on pages 282-83. 
The first column gives the distance of the exploring coils from the cen¬ 
ters of the magnets at the end of each step. The second and fourth 
columns gives the number of C. G. S. lines of induction passing out from 
the magnet at 14° and 99.°5 C. respectively in the step of the coil shown in 
the first column. The third column gives the change in distribution. The 
fifth and sixth columns gives the magnetization at the two tempera¬ 
tures, i. e. the number of C. G. S. lines of induction per square centime¬ 
ter passing through the magnet at the point indicated. The seventh 
column gives the proportional change in magnetization at different 
points of the magnet. 
In the fifth and sixth columns two values are given for the center of 
the magnet, calculated from the two ends, and serve to indicate the de¬ 
gree of accuracy attained. The variation is considerably less than one 
per cent. The second and third columns from which all the others are 
calculated give the means of at least five or six separate determinations, 
which agree well among themselves. The results were further checked 
by slipping the coils from the middle of the magnet clear off each end 
at both temperatures. The variation between this measurement and 
the others was always less than one-half of one per cent. This was con¬ 
sidered quite good as it is impossible to slip the coils over this whole 
distance at the same rate at which they were slipped over the small 
divisions. 
In plate XI our results are shown graphically. The full lines give the 
distribution at 14° C and 99.°5 C. The dotted lines give the change in 
distribution due to this change in temperature. The scale of ordinates 
in the last curves is ten times that of the distribution curves. 
In addition to the checks upon accuracy already mentioned, the fol¬ 
lowing entirely independent determinations of the magnetic moments 
of the two magnets were made. In the tables of distribution pp. 282-83, we 
are given the number n of lines of induction issuing from little divis¬ 
ions of the bar, as well as the distance d of these divisions from the 
center of the magnet. A first approximation to the moment is given 
by the formula: 
M=4 ~ 2 nd. 
47T 
