280 STAFF COMMANDER EVANS AND MR. A. SMITH ON THE MAGNETIC 
a C of 19°, which was caused by the ship having been built, plated, and moored with the 
starboard side South, is reduced to +‘037 or 2° 10' by lying for six months with the 
port side South. This amount does not alter materially while the ship is allowed to 
swing, but when she is twenty* five days in dock with the starboard side South, it suddenly 
rises to + T23 or 7°. 
SB, it will be observed, changes much less at first, and hardly changes at all afterwards ; 
this difference must be attributed in part to this, that while the whole of (5 is to be 
attributed to subpermanent magnetism arising from horizontal induction in transverse 
hard iron, a large part of the original S3 was probably caused by the transient magnetism 
arising from vertical induction in soft iron, and a further part by the subpermanent mag- 
netism arising from vertical induction in hard iron, so that possibly not more than TOO was 
caused by the subpermanent magnetism arising from induction from the headward com- 
ponent of the horizontal force, nearly the whole of which may have been removed by six 
months’ reversal of her direction, so as to leave little room for subsequent change of S3. 
In connexion with this part of the subject we may observe that the same circumstances 
which cause the transient magnetism arising from horizontal induction in transverse 
iron (—&) to be greater than the transient magnetism arising from horizontal induction 
in fore-and-aft iron ( — a), lead us to expect that the subpermanent magnetism arising 
from horizontal induction in transverse hard iron ((5) will be greater than the subper- 
manent magnetism arising from horizontal induction in fore-and-aft hard iron (changing 
part of S3), and that consequently we should expect the relative changes of 6 which take 
place on a change of direction to be greater than those of S3, and this will be found to 
be verified in almost all cases, except when the ship has been built nearly North and South. 
3. After a certain time, which may be roughly estimated at a year after launching, 
this process seems to stop, and the values of B and G remain remarkably permanent. 
The former paper* contains numerous examples of this in ordinary iron-built ships. 
This will appear also from the following instances of the iron-plated ships. 
Standard Compass. 
a <£. 
Warrior. 
September 1861 . . . 
-•449 
-•124 
October 1861 . . . . 
-•409 
-•092 
July 1862 .... 
-•321 
-•114 
June 1863 .... 
—-■317 
— T32 
July 1864 .... 
-•311 
-•054 
October 1864 .... 
-•307 
-•072 
Defence. 
February 1862 .... 
+•464 
+ •005 
March 1863 .... 
+ •379 
-•034 
December 1863 . . , . 
+ •403 
-•016 
April 1864 .... 
+ •391 
f 
© 
© 
-a 
October 1864 .... 
+ •379 
-•034 
* Philosophical Transactions, Part II. 
I860. 
