will; sometimes it will not. But if it will, no one has a right to
ask you to try any other explanation.

Suppose, for instance, that you found a dead bird on the top of a
cathedral tower, and were asked how you thought it had got there.
You would say, "Of course, it died up here." But if a friend said,
"Not so; it dropped from a balloon, or from the clouds;" and told you
the prettiest tale of how the bird came to so strange an end, you
would answer, "No, no; I must reason from what I know. I know that
birds haunt the cathedral tower; I know that birds die; and
therefore, let your story be as pretty as it may, my common sense
bids me take the simplest explanation, and say--it died here." In
saying that, you would be talking scientifically. You would have
made a fair and sufficient induction (as it is called) from the facts
about birds' habits and birds' deaths which you know.

But suppose that when you took the bird up you found that it was
neither a jackdaw, nor a sparrow nor a swallow, as you expected, but
a humming-bird. Then you would be adrift again. The fact of it
being a humming-bird would be a new fact which you had not taken into
account, and for which your old explanation was not sufficient; and
you would have to try a new induction--to use your common sense
afresh--saying, "I have not to explain merely how a dead bird got
here, but how a dead humming-bird."

And now, if your imaginative friend chimed in triumphantly with: "Do
you not see that I was right after all? Do you not see that it fell
from the clouds? that it was swept away hither, all the way from
South America, by some south-westerly storm, and wearied out at last,
dropped here to find rest, as in a sacred-place?" what would you