is found true on the application of all available tests in an overwhelming proportion of the extant MSS in which ancient literature has been preserved.
9. This general proposition respecting the average results of transcription requires to be at once qualified and extended by the statement of certain more limited conditions of transmission with which the New Testament is specially though by no means exclusively concerned. Their full bearing will not be apparent till they have been explained in some detail further on, but for the sake of clearness they must be mentioned here.
10. The act of transcription may under different circumstances involve different processes. In strictness it is the exact reproduction of a given series of words in a given order. Where this purpose is distinctly recognised or assumed, there can be no errors but those of workmanship, 'clerical errors', as they are called; and by sedulous cultivation, under the pressure of religious, literary, or professional motives, a high standard of immunity from even clerical errors has at times been attained. On the other hand, pure clerical errors, that is, mechanical confusions of ear or eye alone, pass imperceptibly into errors due to unconscious mental action, as any one may ascertain by registering and analysing his own mistakes in transcription; so that it is quite possible to intend nothing but faithful transcription, and yet to introduce changes due to interpretation of sense. Now, as these hidden intrusions of mental action are specially capable of being restrained by conscious vigilance, so on the other hand they are liable to multiply spontaneously where there is no distinct perception that a transcriber's duty is to transcribe and nothing more; and this perception is rarer and more dependent on
