But the unity of this source is much harder to demonstrate than is the unity of Q. A considerable amount of the material, aside from the Q material, in these sections is more or less closely duplicated by Matthew, and the Perean source or its equivalent in parts must therefore have been used by him also. Matthew's demonstrated faithfulness to his sources raises serious doubt as to whether he could have known this Perean source and have omitted so much of it. The assumption that he did so, and the assignment of the double tradition thruout this portion of Luke, would require also an entire rearrangement of Q. Burton accepts this requirement, and, instead of Q, goes back to the Logia as a special source of Matthew. The fact that some of this material in the so-called Perean section of Luke may easily be assigned to his own invention, and that in the larger part of it where he is not duplicated by Matthew his own hand can be clearly seen in additions and rearrangements, would seem to tell against the unity of the Perean source, or against the assumption of any Perean source properly so called, and common to Matthew and Luke. On the whole the hypothesis of a Perean source does not seem to the writer to have been substantiated. OTHER POSSIBLE SOURCES FOR MATERIAL PECULIAR TO LUKE Suggestion has been made in connection with a few of the passages considered on pp. 193-206 as to a possible Jerusalem source. Nothing can perhaps be said in support of such a hypothesis, except what is suggested in the analysis on those pages and lies upon the surface of the passages. Another possible clew to the determination of one of Luke's sources lies in the material that has to do particularly with women. Compare the raising of the widow's son and the speech of Jesus referring to the Old Testament widow; the ministering women, Mary and Martha, and the speech of the woman about the mother of Jesus. The writer does not consider this (or the preceding) to be anything more than a suggestion. CONCLUSIONS REGARDING Q MATERIAL IN THE SINGLE TRADITIONS OF MATTHEW AND LUKE The preceding investigations represent the recension of Q used by Matthew as containing about two hundred and sixty-seven verses, or parts of verses. Of these ninety-eight are so closely parallel to Luke as to be marked simply Q. Eighty-nine, paralleled in Luke, but with divergences such as to indicate a different wording in the source that lay before Matthew and Luke and eighty without any parallels in Luke, are assigned to QMt. The recension of Q used by Luke, according to our analysis, contained about two hundred and thirty-eight verses or parts of verses. Of these, ninety-four are closely enough paralleled in Matthew to be assigned simply to Q; eighty-one are paralleled in Matthew, but with such differences as to suggest different wording in the source; and sixty-three are peculiar to Luke.