Page:Popular Science Monthly Volume 82.djvu/230

226

of the botanical historian and has been carried further than that of any other plant that has been cultivated since before recorded time and of which the wild prototype is unknown.

The clues upon which the botanical detective works are many, and it is only by dovetailing numerous facts that the probability of a correct conclusion is increased until it is beyond question. That criminal detectives can establish a reasonable proof by circumstantial evidence was shown long ago by Poe. Mathematics does not recognize a series of coincidences. Coincidences do occur but not in series. If a series of facts point to the same conclusion, the probability that that conclusion is correct increases by multiplication, not by addition. If the probability that one throws heads with a coin is one half, the probability that he throws a pair of heads with two coins is one half times one half, or one quarter, not one half plus one half which would be a certainty. Thus the fictitious reasoners of Poe and Doyle have argued that if a series of independent circumstances point in a single direction, that direction is the proper one. If certain facts seem to be outstanding, they must be looked to, for their fallacies will sooner or later come to light. The same is true in botanical history as the following incident shows: