There must be no speculation upon a man's guilt. A man is regarded as innocent so long as it is impossible to connect to him the last link in a long chain of circumstantial evidence. In the brief accounts of various celebrated trials in the following pages an attempt has been made to give an outline of the scientific circumstantial evidence that has led to the conviction or acquittal of the prisoners. In some of these trials proof of guilt has been overwhelming, although the testimony of an eye-witness has been lacking, but in others the Scotch verdict of "Not proven" (a curious equivalent of which, however, was once given in the trial of Mrs. Rudd) would be a more fitting deduction from the evidence, than the alternative of "Guilty" or "Not guilty," which is all that is allowed by the English law. A good illustration of the value of scientific proof was seen, in 1884, at the trial of a woman named Gibbons on the charge of having shot her husband. For the defence it was urged that the man had committed suicide. There were four bullet wounds from a revolver in the body, and the medical evidence went to prove that although any one of the wounds might have been inflicted by the man himself, it was extremely improbable that all of them had been. Moreover, some of them were in such a position that they could only have been self-inflicted if the revolver had been held in the left hand, whereas witnesses testified that the deceased was not left-handed. The prisoner was found guilty. Attempts have frequently been made by defending counsel to obtain permission for a scientific man to be present on behalf of a prisoner at any examination made before a trial, but all such requests are invariably refused. It is quite a common occurrence, however, for the evidence given by scientific witnesses for the prosecution to be controverted by scientific witnesses for the defence, and the most recent instance of the kind at the trial of Crippen will be fresh in the memory of everyone. Where there is any possibility of doubt it should be possible for every prisoner to obtain scientific assistance. An accused person who lacks the means to procure legal assistance in his defence has assigned to him by the Court a barrister who will represent his interests and see that they do not suffer from ignorance of legal technicalities. This principle might well be extended so as to cover the ground of scientific evidence. Under the present conditions the prosecution has unlimited facilities for applying every description of test, but it has not always been easy for the representatives of the accused person to obtain scientific help in criticising the nature of this evidence. Scientific evidence should be, and usually is, quite impartial, but the everyday conflict of honest opinion in civil actions illustrates the possibility of mistakes occurring or of certain points that would tell in favour of the accused being overlooked.