gives no assistance in determining the order in which any two occurrences will take place. Thus, if it be shown that an event will occur once in a million of times, in the first million it may be the last in the series, and in the second it may be the first; and that will bring them side by side. Many years ago there was a famous lawsuit in New England. A wealthy woman died, leaving large sums for benevolent purposes, and to her niece—already very rich—almost a million of dollars. The niece made strenuous efforts to break the will. A codicil was produced, the signature of which was found to be exactly like another signature of the testatrix. It was hinted, if not explicitly charged, by the counsel for the will that it was a forgery. Professor Peirce of Harvard University was brought in as a witness. He testified that not more than once in many millions of times would two signatures of the same person be written precisely alike. From this it was designed to raise the presumption that where there is an exact coincidence it must have been done by tracing. The court sustained the will on other grounds, and declined to decide that question. But the force of a presumption of this kind is much weakened, if not destroyed, by the fact that all to which Professor Peirce testified might be true, yet the two similar signatures might occur in the same month. Mr. Proctor states it thus:
The business of life insurance can be carried on with certainty, provided the system be constructed upon averages deduced from a sufficiently large num-
