MR. J. S. TOWNSEND ON MAGNETIZATION OF LIQUIDS. 
547 
iron per cub. centim. gave values of Jc the same as if water alone were present, so 
that when iron appears in the acid radical it is at least 100 times less magnetic than 
when it takes the part of a metal in a salt. 
Effect of Temperature. 
In estimating the effect of temperature it is necessary to use a tube of about one- 
third the sectional area of that on which the solenoid is wound, so as to be able to 
prevent the latter from being heated by radiation and convection. 
The time required to keep the hot tube inside the solenoid while adjusting is about 
half- a-minute, so that it is an easy matter to prevent the coils getting heated. The 
efficiency of the method employed was proved lay the balance returning to its original 
zero when the liquid was taken out and the reduction in the turns in c restored. 
Uniformity of temperature along the column of the liquid is secured by inverting the 
tube containing it before inserting it into the solenoid. 
Fig. 3. 
The results of these experiments are best shown by drawing curves, fig. 3, the x 
ordinates denoting temperature centigrade, the y ordinates being proportional to k, a 
length corresponding to 50 being subtracted from the latter for convenience of 
representation. 
The first curve was got from experiments done on a solution of ferric chloride 
containing ’086 grm. of iron per cub. centim., the 2nd from ferrous chloride 
containing T48 grm. of iron per cub. centim., the 3rd from ferric sulphate containing 
4 a 2 
