similarity_ which was the very opposite to the marks of healthy human
news.

I would particularly insist upon this last point. It is little
understood and it is vital.

If we want to know what to think of a fire which has taken place many
miles away, but which affects property of our own, we listen to the
accounts of dozens of men. We rapidly and instinctively differentiate
between these accounts according to the characters of the witnesses.
Equally instinctively, we counter-test these accounts by the inherent
probabilities of the situation.

An honest and sober man tells us that the roof of the house fell in.
An imaginative fool, who is also a swindler, assures us that he later
saw the roof standing. We remember that the roof was of iron girders
covered with wood, and draw this conclusion: That the framework still
stands, but that the healing fell through in a mass of blazing
rubbish. Our common sense and our knowledge of the situation incline
us rather to the bad than to the good witness, and we are right. But
the Press cannot of its nature give a great number of separate
testimonies. These would take too long to collect, and would be too
expensive to collect. Still less is it able to deliver the weight of
each. It, therefore, presents us, even at its best when the testimony
is not tainted, no more than one crude affirmation. This one relation
is, as I have said, further propagated unanimously and with extreme
rapidity. Instead of an organic impression formed at leisure in the
comparison of many human sources, the reader obtains a mechanical one.
At the same moment myriads of other men receive this same impression.
Their adherence to it corroborates his own. Even therefore when the