For the most part, and especially in these latter ages, persons are culled out for this work from amongst the common people, upon the sole consideration of well-speaking, as if we were to learn grammar from them; and the men so chosen have fair reason, being hired for no other end and pretending to nothing but babble, not to be very solicitous of any part but that, and so, with a fine jingle of words, prepare us a pretty contexture of reports they pick up in the streets. The only good histories are those that have been written themselves who held command in the affairs whereof they write, or who participated in the conduct of them, or, at least, who have had the conduct of others of the same nature. Such are almost all the Greek and Roman histories: for, several eye-witnesses having written of the same subject, in the time when grandeur and learning commonly met in the same person, if there happen to be an error, it must of necessity be a very slight one, and upon a very doubtful incident. What can a man expect from a physician who writes of war, or from a mere scholar, treating of the designs of princes? If we could take notice how scrupulous the Romans were in this, there would need but this example: Asinius Pollio found in the histories of Caesar himself something misreported, a mistake occasioned; either by reason he could not have his eye in all parts of his army at once and had given credit to some individual persons who had not delivered him a very true account; or else, for not having had too perfect notice given him by his lieutenants of what they had done in his absence.--[Suetonius, Life of Caesar, c. 56.]--By which we may see, whether the inquisition after truth be not very delicate, when a man cannot believe the report of a battle from the knowledge of him who there commanded, nor from the soldiers who were engaged in it, unless, after the method of a judicial inquiry, the witnesses be confronted and objections considered upon the proof of the least detail of every incident. In good earnest the knowledge we have of our own affairs, is much more obscure: but that has been sufficiently handled by Bodin, and according to my own sentiment --[In the work by jean Bodin, entitled "Methodus ad facilem historiarum cognitionem." 1566.]--A little to aid the weakness of my memory (so extreme that it has happened to me more than once, to take books again into my hand as new and unseen, that I had carefully read over a few years before, and scribbled with my notes) I have adopted a custom of late, to note at the end of every book (that is, of those I never intend to read again) the time when I made an end on't, and the judgment I had made of it, to the end that this might, at least, represent to me the character and general idea I had conceived of the author in reading it; and I will here transcribe some of those annotations.