Here the line of the narrative is again broken abruptly. Joseph is never mentioned again. 21'- 22- re-tells Mt 2"-, with elaborations. The magi have seen ' a star of enormous size, shining among these stars and eclipsing their light.' The star conducts them to the cave, where the magi .see ' the infant with his mother Mary ; and they brought out of their wallet gifts of gold, incense, and myrrh. And being instructed by the angel not to enter Judaea, they went to their own land by another road.' § The omission of Joseph would not of itself be significant (in view of Mt 2'-'2), were it not that in 22^'^ the initiative is assigned to Mary instead of to Joseph (as in Mt 2'*'-). Hearing of Herod's order to massacre all children of two years and under, Mary hides the child Jesus in an ox- stall. Evidently, the original narrative ignored the flight to Egypt. But what it substituted for this remains a mystery, for at this point (22^) the story suddenly breaks into an account of John the Baptist and his parents. The child John is among the infants sought for by Herod, and Elizabeth in despair prays to a mountain in the hill-country, 'O mountain of God, receive mother and child.' The mountain immediately parts in two and shelters them, protected by a light ('for an angel of the Lord was with them, watching over them'). Herod, unable to make Zechariah (who is high priest) confess the whereabouts of his child, has him murdered inside the Temple, on the ground that • his son is to be king over Israel,' At day- break, as Zechariah does not come out, one of the priests ventures inside ; he sees clotted blood beside the altar, and hears a voice saying, ' Zechariah has been murdered, and his blood shall not be wiped up until his avenger comes.' His body is never found, but his blood turned to stone. The Simeon of Lk 2"* is chosen by lot to succeed him, and with this the story ends. The epilogue runs : ' I, James, the writer of this history, when a riot arose in Jerusa- lem at the death of Herod, withdrew myself to the desert till the riot in Jerusalem ceased, glorifying the Lord God who gave me the gift and the wisdom to write this history.' The book thus professes to be written not only by an eye-witness but imme- diately after the event.
In spite of Zahn's and Conrady's arguments to
• The Syriac fragment passes straight from 182 to 19^.
t Possibly echoed in Clem. Strom, vii. 16. 93.
i Jesus, in the Syriac as in pseudo-Matthew (see below, p. 488).
§ The simplicity of the story is noticeable ; in the primitive form (expanded in the versions and later MSS) the magi do not even adure the child, and no attempt is made to name them, as in the Armenian version, which calls them Melchior, prince of Persia, Baltasar, prince of India, and Caspar, prince of Arabia. The angel goes to them at once after the Annunciation, ' and they were led by the star for nine months, and then came and arrived in time for the birth from the holy virgin.' This is reproduced in the Coventry Nativity play.
GOSPELS (UNCAN0:N'ICAL)
GOSPELS (UNCA^ONICAL) 485
the contrary, it is almost necessary to postulate the composite character of the Protevangelium, although the sources cannot be disentangled with much precision. Even in 1-18^ there are traces of different strata, e.g. the sudden introduction of Joseph in 9\ and the episode of Mary sewing the purple and scarlet* for the veil of the Temple (10, 12). The latter episode could be parted from the context not only without difficulty but with a gain to the sequence of the narrative.! On the other hand, neither 1-18^ nor 18^-22^ can be regarded as complete sources. The legend of Zechariah's murder in 22*- 24, on the other hand, is a water- mark of late origin. In the light of the investiga- tions by A. Berendts.J it is clearly subsequent to Origen, who knows quite a different version of Zechariah's death — one which connects it closely with the virginity of Mary (he was murdered, according to this tradition, between the Temple and the altar, for having permitted Mary to enter the court of the virgins after she had given birth to Jesus). Had Origen read 22^-24 in his j8i'/3\os 'Ia/cw/3ou, he would not have written as he has done upon Mt 23*^ For the existence of the legend in the form of 22*-24 the first evidence is from Peter of Alexandria (t A.D. 311), and even this evidence is not absolutely decisive.
Whether the composite work underwent suc- cessive expansions or, as is less likely, was recast by a Gnostic author, P-18^ which is practically a yivv7iffi,s Ma/jiay, probably belonged to the book of James, from which Origen quotes. His quotation is based on this part, and on this part alone ; the rest of the book never mentions the other children of Joseph. \i the conclusion (25) was part of the original romance, the story must have included the incidents of Herod's massaci-e, tliough in a form difieriiig from that preserved in the Apocalypse of Zechariah § as it now appears in 22^-24. For some reason, the latter must have been substituted for the original conclusion, or added to a narrative which had lost its ending. Whether 18^-21^ was also an extract from some Apocryphum Josephi, which became appended to 1-18\ or whether the author of the book of James himself combined the fragment with his other source, is a problem which cannot be decided definitely either way, in view of the obscurity surrounding the literary origins of this as of most other pseudepigrapha.