in the East, slept in the open air. These would, of course, be driven into Jerusalem by the approach of the Roman armies, and thus the numbers at the beginning of the siege might have amounted to a million or thereabouts. But the notion of nearly three millions being crowded into the area above described is simply preposterous. But if Josephus’s statistics on these two important points are to be rejected as wholly untrustworthy, how are we to credit his assertions in matters of very nearly the same kind? He tells us that Galilee in his time contained more than two hundred towns and villages, no one of which held less than 15,000 inhabitants.[240] If this were indeed the case, that province, scarcely larger than one of the largest of our English counties, must have had a population of fully three millions, while that of the whole of Palestine would approach ten millions. Few readers will be found to credit this. At the same time more than one trustworthy writer affirms that Palestine was a thickly populated country. The population to the square mile is said to have been larger in it than in any other portion of the Roman dominions. Diodorus,[241] Strabo,[242] Tacitus,[243] and Dion Cassius[244] all concur in this; and therefore, though we cannot accept Josephus’s statements as being even approximately accurate, they may be admitted so far, as establishing the numerous population of Palestine at the time of the siege. Nor are we wholly without means of forming an estimate as to its amount, independently altogether of the above-named writers. Thus Hecatæus of Abdera (quoted by Joseph. Ap. i. 21) says that Jerusalem in his time (A.D. 312) contained 120,000 inhabitants. Presuming the average increase of population to have taken place, according to this reckoning, Jerusalem at the time of the siege would contain about 600,000—agreeing closely with Tacitus’s estimate. According to Maccab. II., the city at the date of Antiochus Epiphanes, A.D. 180, had 160,000, or, according to others, 180,000. This would make the number of residents at the outbreak of the civil war somewhat less; but there would be no material difference. On the whole, we may assume that, by dividing Josephus’s estimates by three, we approximate to the real number. According to this, the census of the Holy Land, A.D. 71, would be about three and a half millions, and the total of persons besieged in the Holy City something under one million.