But supposing it to be well understood at the moment, the exact wording of it can rarely be recalled, unless the witness's memory were tantamount in minuteness and accuracy to the record of a short-hand writer. He is consequently permitted to give an abstract, or, as it is usually called, the substance of what occurred. But here a new difficulty arises; to abstract correctly is an intellectual effort of no mean order, and is rarely accomplished with a decent approach to perfection. Let the juryman bear this in mind. He will be often tempted to rely on alleged confessions of prisoners sworn to by witnesses who certainly desire to speak the truth. These confessions often go so straight to the point, that they offer to the juryman a species of relief from that state of doubt, which, to minds unpracticed in weighing probabilities, is irksome, almost beyond description. Speaking from the experience of thirty years, I should pronounce the evidence of words to be so dangerous in its nature as to demand the utmost vigilance, in all cases, before it is allowed to influence the verdict to any important extent. While I am on the subject of evidence, infirm in its nature, I must not pass over that of identity of person. The number of persons who resemble each other is not inconsiderable in itself; but the number is very large of persons, who, though very distinguishable when standing side by side, are yet sufficiently alike to deceive those who are without the means of immediate comparison.