In about an hour's time I took another set of observations with the utmost care; the mean result of these was about the same as that of the first set.

I asked myself why, with my boasted self-dependence, I had not done at least better than this.

Then I went in search of a discrepancy in the tables, and I found it.

In the tables I found that the column of figures from which I had got an important logarithm was in error.

It was a matter I could prove beyond a doubt, and it made the difference as already stated.