252
POLE
ON
MENTAL
CALCULATION.
[Selected
was
it
a matter of simple memory, for
it
would be equally
impossible for any one to recollect the logarithms, to eight places,
of
hundreds of thousands
of
numbers.
It
was, however, founded
on
a power of memory, of a more
limited range. Mr. Bidder's process may be said briefly
to
be
this
:
having stored in his memory the logarithms of a few
simple numbers, he was able, by his wonderful skill in dealing
with figures, to make use of them, mentally, to calculate accurately
the logarithms of any other numbers however large.
A
brief
description will give an idea of the manner in which this was
done.
In
the
first
place, as to the mnemonic bases used. He knew by
heart the logarithms of a great many
small prime numbers,
nearly
all, he said, under
100,
and some few above.
Secondly, it will be easily understood that these would enable
him
to
calculate the logarithms of any large numbers which were
nmltiples of those he knew. Thus for the logarithm of 63 he had
only to add together the logarithms of
7
and
9
;
for the logarithm
of 3567, he had only to sum up the logarithms of
29,
41,
and 3,
and
so
on.
In
carrying out this procesg he had an almost miraculous power
of
seeing, as it were intuitively, what factors would divide any
large given number not a prime. Thus if he was given the
number
17,861,
he would instantly remark
it
was
=
337
X
63
;
or
he
would see
as
quickly that
1659
was
=
79
X
7
X
3.
He could
not, he said, explain how he did this,
it
seemed like a natural
instinct to him.
These two qualifications, therefore-the knowledge of the logar-
ithms of certain small primes, and the power of reducing large
compound numbers to their component factors-constituted the
foundation
on
which Mr. Bidder's operations proceeded. They
would suffice for determining the logarithms of a great many
numbers by simply adding together the logarithms of their com-
ponent factors, which he could
do
mentally with the greatest ease.
But
it
still remains to be explained how he treated the case of
large primes, and this is really the great interest of his process.
His first endeavour was to find some multiple number
cery near
the number given, differing from
it
only
in
the last place of
figures, and if possible only by unity. For example, being given
the number
1051,
he would easily find the logarithm
of
1050
(=
30
X
7
X
5)
to which he would then have to make an addition
for the difference
of
1.
This addition would be very important
(for the accuracy aimed at), as
it
would affect the last five figures