vi. 29; compare Matt. v. 39, 40), [Greek:  ti me legeis agathon] and [Greek:  oudeis agathos ei mae] (Luke xviii. 19; compare Matt. xix. 17), [Greek:  meta tauta mae echonton] ([Greek:  dunamenous] Justin) [Greek:  perissoteron] (om.  Justin) [Greek:  ti poiaesae k.t.l.] (Luke xii. 4, 5; compare Matt.  X. 28), [Greek:  paeganon] and [Greek:  agapaen tou Theou] (Luke xi. 42).  In the parallel passage to Luke ix. 22 (=Matt xvi. 21= Mark viii. 31) Justin has the striking word [Greek:  apodokimasthaenai], with Mark and Luke against Matthew, and [Greek:  hupo] with Mark against the [Greek:  apo] of the two other Synoptics.  This last coincidence can perhaps hardly be pressed, as [Greek:  hupo] would be the more natural word to use.

In the cases where we have only the double synopsis to compare with Justin, we have no certain test to distinguish between the primary and secondary features in the text of the Gospels.  We cannot say with confidence what belonged to the original document and what to the later editor who reduced it to its present form.  In these cases therefore it is possible that when Justin has a detail that is found in St. Matthew and wanting in St. Luke, or found in St. Luke and wanting in St. Matthew, he is still not quoting directly from either of those Gospels, but from the common document on which they are based.  The triple synopsis however furnishes such a criterion.  It enables us to see what was the original text and how any single Evangelist has diverged from it.  Thus in the two instances quoted at the beginning of the last paragraph it is evident that the Lucan text represents a deviation from the original, and that deviation Justin has reproduced.  The word [Greek:  isangeloi] may be taken as a crucial case.  Both the other Synoptics have simply [Greek hos angeloi], and this may be set down as undoubtedly the reading of the original; the form [Greek:  isangeloi], which occurs nowhere else in the New Testament, and I believe, so far as we know, nowhere else in Greek before this passage [Endnote 128:1], has clearly been coined by the third Evangelist and has been adopted from him by Justin.  So that in a quotation which otherwise presents considerable variation we have what I think must be called the strongest evidence that Justin really had St. Luke’s narrative, either in itself or in some secondary shape, before him.

We are thus brought once more to the old result.  If Justin did not use our Gospels in their present shape as they have come down to us, he used them in a later shape, not in an earlier.  His resemblances to them cannot be accounted for by the supposition that he had access to the materials out of which they were composed, because he reproduces features which by the nature of the case cannot have been present in those originals, but of which we are still able to trace the authorship and the exact point of their insertion.  Our Gospels form a secondary stage in the history of the text, Justin’s quotations a tertiary.  In order to reach the state in which it is found in Justin, the road lies through our Gospels, and not outside them.