that it is difficult to know what to do with them. They are assuredly not clerical errors, but genuine records of pronunciation, whether of the apostolic age or some other early time, and have to a certain extent the support of inscriptions, even of inscriptions from Attica. They seem to be chiefly relics of the digamma, and are interesting as signs of the variety of spoken language which often lies concealed under the artificial uniformity of a literary standard. The range of good MSS supporting them in one place or another is remarkable, and in some few places they can claim a large aggregation of good MSS: yet in others they receive but little attestation, and usually they receive none at all. In two or three cases we have admitted them to the text, content elsewhere to leave them for the present as alternatives in the Appendix, where any needful details as to these or other accessory marks will be found. The amply attested reading οὐκ ἔστηκεν in John viii 44 does not come under the present head, ἔστηκεν being merely the imperfect of στήκω as it appears also to be in Apoc. xii 4. The sense of an imperfect rather than a present is required by the context, which must refer to the primal apostasy as representing the Jews' abandonment of the truth into which they were born; and there is a fitness in the virtually intensive force ('stand fast') which belongs by prevalent though not constant usage to στήκω. The imperfect of this somewhat rare verb is not on record: but imperfects are too closely connected with presents to need separate authority, and multitudes of unique forms of verbs are known only from single passages. The aspiration of αυτοῦ used reflexively is discussed in the Appendix.
408. The breathings of proper names possess a semblance of documentary evidence in the Latin version and its presentation of names with or without H. Yet, however early the first link in the Latin chain may be, it is evidently disconnected from the Palestinian pronunciation of Greek, the true object of search. The serious inconsistencies and improbabilities contained in the Latin usage condemn it equally on internal grounds: it is obviously due rather to unconscious submission to deceptive analogies and associations of sound than to any actual tradition. The breathings of Greek and Latin proper names can usually be fixed by the etymology: where this fails, it is seldom difficult to find direct or indirect authority in coins, inscriptions, or even early MSS of Latin authors. The well