Page:Mind (New Series) Volume 8.djvu/94

 80 A. F. RAVENSHEAE: is available ; only when either absolutely or practically it is- a necessity. A slight development of this shows us the Expert in his proper place he should be employed only in so far as his assistance is unavoidable. He should, to make the strictest use of his powers, be referred to only to prove or to point out unanalysable facts of observation or experiment not without his aid perceptible to or attainable by the in- expert. If this rule be departed from in any given case and the expert asked his opinion on a matter as a whole it should be clearly kept in view that such departure is justifiable only as a concession to convenience. Concatenation of Testimony. It is an obvious conclusion from the preceding discussion that a mere random assertion uttered we know not by whom or under what circum- stances is in general of little value as testimony. State- ments have weight as testimony only in so far as we already have information about the assertor independently of the subject under consideration. Testimony then, in general, consists in assertions whose trustworthiness can be judged through the medium of independent information about the assertor. The case in which this independent information is obtained wholly or in part through the medium of other testimony deserves, on account of its wide occurrence and the import- ance of its uses, to be especially singled out. Our sources of testimony need not be known to us first-hand ; indeed, perforce they are not usually so. By the various ways of obtaining a knowledge of distant facts, including reliance on testimony, we may obtain information about persons or writers at a distance or in the past sufficient to enable us to judge as to the trustworthiness of any statement that can be properly attributed to them. It is of some interest that testimony reached and vouched for in this manner can be shown to be free from the obvious objections to statements that have been handed on from person to person, each relying on his predecessor. The latter certainly bears a superficial resemblance to the use of testimony for establishing the credentials of other witnesses, and is liable to be confused with it. Indeed, in enforcing the importance of the distinction between the " self-infirma- tive chain " and the " self-corroborative chain," Bentham l and those who follow him seem to make no mention of any other possible way in which the testimony of a number of persons might be concatenated. 1 Rationale of Judicial Evidence, vol. iii., p. 224, and chap, x., bk. vi. ; Mill's System of Logic, chap, xxiii., 6.