In mechanics _nothing is stronger than its weakest part_, and it would be well if a similar apothegm could be extended to the criticism of such compositions as the Annales Cambriæ, and the Saxon Chronicle. It would be well if we could say that in chronological tables _nothing was earlier than the latest entry_. In common histories we do this. The common historian is always supposed to have composed his work subsequent to the date of the latest event contained in it--a few exceptions only being made for those authors whose works treat of cotemporary actions. So it is with the annalist whose Annals, more ambitious in form than the bare chronicle, emulate, like those of the great Roman historian, the style of history. But it is not so when the notices pass a certain limit, and become short and scanty. They then suggest a comparison with the parish register, or the Olympic records, and change their character altogether. No longer mere chronological works, emanating from the pen of a single author, and referrible to some single generation, subsequent, in general, to a majority of the events set down in them, they are the productions of a series of writers, each of whom is a registrar of cotemporary events. By this an undue value attaches itself to works which have nothing in common with the register but the form. Now, if genuine traditions are scarce, real registers are scarcer. In both cases, however, the false wears the garb of the true, and, in both cases, writers shew an equal repugnance to scrutiny. This is to be regretted; since with nine out of ten of the chronicles that have come down to us, it is far more certain that their latest facts are earlier in date than the author who records them, than that the earliest possible author can have been cotemporary with the first recorded events. The Anglo-Saxon Chronicle may illustrate this. It ends in the reign of Stephen; yet the writer of even the last page may have been anything but a cotemporary with the events it embodies. It begins with the invasion of Julius Cæsar. A cotemporary entry--the essential element of registration--is out of the question here. The general rule with compositions of the kind in question is, that they fall into two parts, the first of which cannot be of equal antiquity with the events recorded, the second of which may be; and we are only too fortunate when satisfactory proofs of cotemporary composition enable us to convert the possible into the probable, the probable into the certain--the _may_ into the _must_. Even when this is the case, the proportions of the cotemporary to the non-cotemporary statements are generally uncertain--a question of _more_ or _less_, that must be settled by the examination of the particular composition under consideration. Whatever may be the other merits of the _Annales Cambriæ_, it has no claim to the title of a register during the sixth century--and, _a fortiori_ none during the fifth.