narrative proves that it cannot have been winter. You cannot allow, he says, as much as eight oi nine months — April to January — after the pass- over of ch. 2, for there is nothing in the story to suggest 80 long a period, and the impression made on the Galileans at that passover was still fresh in their minds when Christ came on to Galilee after leaN'ing Samaria (4") ; nor can you allow as much again — .Tanuary to October — between this episode and the Feast of Tabernacles soon to be mentioned : * Orig. in Jn. tom. xiii. 39, 41.
It is not possible at this stage to dismiss either explanation as in itself inadmissible. The words of the verse, especially the In, 'still four months,' have, perhaps, a more natural meaning if the harvest w^as actually four months oU'. On the other hand, the immediate context, the promise of the water which should quench all thirst, has been thought to suggest a warmer season than January, the discourses in St. John's Gospel being, it is said, always fitted to their external surroundings. On this view it has been supposed that the rerpinrivov is a proverbial phrase for the interval between seed- time and harvest, oi5x tiMf's X^7fT« standing for t4 Xeydfiei'oi', the regular idiom for a proverb. It is said in answer that no such words are elsewhere preserved ; but phrases of similar meaning, em- phasizing the interval between preparation and fruition, are common in all languages. It is said also that a strict reckoning would make the interval rather six months than four; but the Kabbis (see Wetstein, ad loc.) were accustomed to divide the year into six stages of two months — seed- time, winter, spring, harv'est, summer, dog-days — so that four months does actually cover the period between the two. Considering, too, the ditl'erences of climate in different parts of Palestine, and the differences of season between barley and wheat harvest, there is nothing improbable in supposing that the interval which can be described as one of six months can be described also as one of four.
Origen has really hit the mark in making the relation of the passage to the general chronological arrangement of tlie Gospel the determining factor in a date which could otherwise only be left open. This relation involves, in the first place, a dis- cussion of the third and last of the doubtful time- notices in St. John.
(3) Jn 5'. Alternative readings lopr/j and ^ iopn/j, and alternative explanations of either reading.
il iopT-i) was analj'zed in the discussion of Jn 6* above, and was found to imply either Passover or Tabernacles, though the very existence of a doubt as to the relative precedence of the two feasts ma<le the use of the phrase without further defini- tion unlikely in itself.
ioprii would leave the fea.st intended quite un- certain. Origen and Epiphanius both argue rightly that the indefiniteness excludes Passover ; the former apparently made it Pentecost (as does his follower Cyril, though the text at the head of this section of the Commentary contains the article), the latter gives a choice between Pentecost and Tabernacles (Orig. in Jn. tom. xiii. 39 ; Epiph. B(er. li. 21, Dind. ).t But just as Talicmacles is important enough to rival the claim of Passover to be meant by the definite ii iofuHi, so equally with Passover it is too important to satisfy the in- definite iofnii, which must be referred to one of the less important festivals, Pentecost (May), Trunipeta (September), Dedication (December), or Purim (February).
* The latter part of the arsfument is, of conn*, TitUted \>j Origen's neglect of the Passover of 6^ ; see above.
t The fact that Origen, who certainly did not read the article, uses of the same feast the words «•!/>* tw* ir r») iom rie 'l«i/2a<M> . . . ^twf^ay^JL*tttt (tom. xiil. 64). showi how eealy la oblique references the article would creep in.
CHRONOLOGY OF NEW TEST.
CHRONOLOGY OF NEW TEST. 409
Am between the two readings, the article is found In M C L A 1-118 S3, the Etr.vp. VS3, Eusebius of Cajsarca, Cyril-text (per- hApe, too. Irensjus, since he made the feast a Passover, see above CD Jn Q*) ', it iii omitted by A B D, Ori^en, Epiph. Chrys. uid the Poichal ChronicU, The wei^rht of external evidence favours the latter group, for it has not only early but varied attestation ; whereas the other is of more liomopeneous type, oritnnally purely Alexandrine, and may easily owe its post-Nicene ■upitorters to the influence of Eusebius of Cffisarea, and tbe theory which he brought into prominence of a three years' Ministry with four Passovers. And when to this is added the suspicious character Just shown to attach on internal grounds to i UfiTT, iif^r without the article may confldently claim to repre* lent the text of tbe evangelist.
Thus the first half of the Gospel gives (1) a pass- over, 2"- " ; (2) a note of time, either May or January, 4" ; (3) an unnamed minor feast, 5' ; (4) a second passover, 6*. These could be combined in more than one way to fit into a single year : e.g. (o) Passover — May — any lesser feast — Passover ; or (;3) Passover— January — Purim (February) — Passover.