Sentences in Q, according to Harnack, are generally connected by καί, δὲ being used but seldom. The same is true of Mark. But this only indicates the comparative nearness of both Mark and Q to the Semitic. The same may be said of the preponderance of simple verbs in distinction from compound in both Mark and Q. Ἐὰν is used twice as frequently as εἰ; Mark also appears to use the former thirty-six times and the latter but fifteen. This fact seems to have more significance by reason of the other, that Luke uses one word thirty-two and the other thirty-three times. Matthew, however, uses ἐὰν exactly twice as often as εἰ. When we remember that all we have of Q is contained in Matthew and Luke, and only a small portion of it in Mark, these facts do seem to indicate a preference for ἐὰν over εἰ as between Mark and Q on the one side and Luke on the other, but between Mark and Q on the one side and Matthew on the other no such contrast appears. Mark and Q are here no nearer to each other, or very little, than either of them is to Matthew. The particle τε is never found in Q.[135] It occurs five times in Mt and seven times in Lk, and but once in Mk. Ὡς in temporal clauses seems to be absent; it is also absent from Matthew, while Luke uses it nineteen times and Mark but once. Clauses with γίνομαι, frequent in Matthew and Luke, are absent from Q; they also occur in Mark; but their absence from Q may be due simply to Q's lack of historical matter. Παρὰ and σὺν are absent; the first is used about evenly by Mark and Matthew, and more frequently by Luke; the second, three times by Matthew, five times by Mark, and twenty-four times by Luke. CONCLUSION AS TO MARK'S DEPENDENCE UPON Q These facts do not all point in the same direction. They seem sometimes to indicate a linguistic affinity between Q and Mark, but this affinity usually extends to Matthew also. What seems to be proved by them is that Mark and Q and Matthew all stand nearer to the Semitic than does Luke. But this is only the obverse of the statement that Luke is the best Grecist. It throws no light upon the literary relation of Mark and Q. Such literary relation, in fact, cannot in the strict sense be "proved." It can only be rendered probable, tho perhaps extremely probable, by the unlikelihood that Mark and Q should have fifty verses in common without any literary relationship. Such relationship being assumed, the dependence is on the side of Mark. CHAPTER VII THE ORIGINAL ORDER OF Q