‘Concerning Barn. iv. 14, [Dr. Stanton] says (p. 33) with justice that this is our earliest instance of the citation of a saying of Christ as “scripture.” In the year A.D. 130, the date upon which he rightly fixes for the composition of the _Epistle of Barnabas_, this estimate of the Gospels would have been in the highest degree surprising, since it is not until A.D. 170 that the next examples of such an estimate make their appearance.’ Dr. Schmiedel goes on (1) to have recourse to the accustomed expedient of suggesting that Barnabas is quoting, not from the words of the Gospel which are identical, but from a passage in 4 Ezra which is quite different; and (2) if that expedient fails, to represent the quotation as a ‘winged word,’ though it is expressly introduced by the formula ‘it is written.’ However, it is not of either of these points that I wish to speak, but rather to call attention to what Dr. Schmiedel thinks would be ‘in the highest degree surprising.’ Why so surprising? What substantial ground have we for expecting anything else? In the first place Dr. Schmiedel begins by exaggerating the significance of the phrase ‘it is written,’ as though on its first extant occurrence it would necessarily imply full canonical authority. And then he goes on to lay stress upon what is really little more than the absence of literature. If we take the whole extant Christian literature between the years 130 and 170 A.D., it would not fill more than a thin octavo volume, and by far the greater part of that is taken up with external controversy. What sort of argument can be drawn from such a state of things as to the exact estimate which Christians formed of their own sacred books? No valid argument can be drawn from it either way, and it is far better simply to confess our ignorance. It is reasonable to suppose that there was a gradual development in the process by which the Gospels attained to the position that we call canonical; but the data to which we have access do not allow us to map out its stages with any precision. It seems to me to be a fundamental defect in the reasoning of Dr. Schmiedel and his school that they fail to see that the real question is, not simply, What is the evidence for this or that proposition? but, What is the relation which the extant evidence bears to the whole body of that which once existed, and how far can we trust the inferences drawn from it?