486 Mr. Babbage and Mr. Herschel's account of the 
27. This principle will, if we mistake not, be found to 
afford at least a plausible explanation of most, if not all the 
phenomena above described, without the necessity of calling 
in any additional hypothesis, or new doctrine in magnetism. 
For the other principle we shall have occasion to employ, 
that magnetic bodies differ exceedingly, both in susceptibility 
of this quality and in the degree of the pertinacity with which 
they retain it (which may be called their retentive power), is 
not an hypothesis, but an acknowledged fact. It is only in 
the mode of its extension to new cases of magnetics that we 
can be led into any fallacies. Whether these two qualities 
(susceptibility and retentive power) be, or be not mutually 
dependent, this is not the place to enquire. Probably they 
are not so, at least directly : and the new facts almost con- 
vert this probability into certainty ; at all events, at present 
we shall for greater generality suppose them independent. 
observed the following facts, which set the principle stated in the text in a very 
clear light. A natural magnet, armed with soft iron, terminating in a cylindrical 
surface, was made to support a load hooked on to another piece of soft iron, termi- 
nating also in a cylindrical surface, so that its contact with the former was limited 
to a physical line. The weight it would usually support varied from 27 to 33 lbs., 
according to the caution used in increasing the load. On loading it with 30 lbs. it 
became necessary to add the remaining weight by degrees, a quarter of a pound at 
a time, and to wait a short time after each addition. At about 32^ lbs. the weights 
usually fell from the magnet, and it was observed on replacing them, that it would 
no longer support more than 30, and that some minutes must be suffered to elapse 
before it could be brought to its former load. It was thus evident that the magnet 
required time for the developement of its full virtue. Again, having loaded the 
magnet by degrees up to 32 lbs., if the contact was broken for an instant by sei2ing 
the iron to which the load was suspended with both hands, detaching it suddenly 
and instantly restoring It, the magnet now continued to suspend 32 lbs., though, 
had the separation been of longer duration, 30 only would have been suspended. 
Time therefore is required to lose, as well as to gain magnetism. 
