[30] Much weight has been attached to a very extensive reproduction of the original text. More important statements are always adduced verbatim; less important, or too cumbersome ones, are abridged, but keeping as closely to the original words as possible. After the sense of a statement has been settled more or less reliably, two cases must be discriminated. If the statement is purely a record of facts, and, still better, if it is exemplified by concrete instances, there is generally no reason to disbelieve it, especially if in the general character of the author there is a guarantee of his trustworthiness; and if he actually has had good opportunities of observing the natives. But if the statement involves a judgment, a generalization, or abstraction, we must be much more careful. Broadly speaking, statements of this latter kind are generally much more contradictory than mere statements of fact. It will be seen that the information concerning the treatment of women by their husbands, concerning sexual matters, and concerning the authority of husbands, will present many more discrepancies than the information concerning the modes of obtaining wives, economics, and other concrete questions. The first category implies much more abstraction and qualifying judgment than the second. It must be borne in mind that statements of the first category are the result of a long and complicated series of mental processes, and that their quality and value is dependent upon many conditions. All these conditions must be mentally analyzed and each of them must be taken into account in order to ascertain its bearing upon the final form in which we find the statement. The conditions in question may be shortly set forth as follows: Did the author possess all the qualities necessary for a good ethnographer? Had he good opportunities to observe the natives and a good method of doing so? Were the latter still in a primitive condition, or in an advanced state of decay? A few words may be said in the first place about this last point.