An excellent illustration of such a case is that of Nan Patterson. She is commonly supposed to have attended, upon the night of her acquittal, a banquet at which one of her lawyers toasted her as "the guilty girl who beat the case." Whether she was guilty or not, there is a general impression that she murdered Caesar Young. Yet the writer, who was present throughout the trial, felt at the conclusion of the case that there was a fairly reasonable doubt of her guilt. Even so, the jury disagreed, although the case is usually referred to as an acquittal and a monument to the sentimentality of juries. The acquittal of Roland B. Molineux is also recalled as a case where a man, previously proved guilty, managed to escape. The writer, who was then an assistant district attorney, made a careful study of the evidence at the time, and feels confident that the great majority of the legal profession would agree with him in the opinion that the Court of Appeals had no choice but to reverse the defendant's first conviction on account of the most prejudicial error committed at the trial, and that the jury who acquitted him upon the second occasion had equally no choice when the case was presented with a proper regard to the rules of evidence and procedure. Indeed, on the second trial the evidence pointed almost as convincingly toward another person as toward the defendant. I have mentioned the Patterson, Thaw, and Molineux trials because they are cases commonly referred to in support of the general contention that the jury system is a failure. But I am inclined to believe that any single judge, bench of judges, or board of commissioners would have reached the same result as the juries did in these instances. It is quite true that juries, for rather obvious reasons, are more apt to acquit in murder cases than in others. In the first place, save where the defendant obviously belongs to the vicious criminal class, a jury finds it somewhat difficult to believe, unless overwhelming motive be shown, that he could have deliberately taken another's life. Thus, with sound reason, they give great weight to the plea of self-defence which the accused urges upon them. He is generally the only witness. His story has to be disproved by circumstantial evidence, if indeed there be any. Frequently it stands alone as the only account of the homicide. Thus murder cases are almost always weaker than others, since the chief witness has been removed by death; while at the same time the nature of the punishment leads the jury unconsciously to require a higher degree of proof than in cases where the consequences are less abhorrent. All this is quite natural and inevitable. Moreover, homicide cases as a rule are better defended than others, a fact which undoubtedly affects the result. These considerations apply to all trials for homicide, notorious or otherwise, the results of which in New York County for ten years are set forth in the following table: