nights, for the latter part of the period between the Creation and the birth of Christ, does not square so exactly with the true place of the third of April in the year of the Nativity; as the former part squared with the assumption of its place in the year of the Exodus: perhaps the following considerations may contribute to mitigate this difficulty, if they do not remove it altogether.

The calculation, for each of the periods in question, proceeded upon the supposition, that the succession of days and nights, between the Creation and the Exodus, and between the Exodus and the Nativity, went on alike; and that the mean length of one νυχθήμερον was always the same with the mean length of another: a supposition which, with respect to the first of the intervals so determined, there is no reason whatever to consider doubtful. But with respect to the latter ; there were two occasions, one in the time of Joshua, B. C. 1520, the other in the reign of Hezekiah a, B. C. 710, when the constant, unvaried, and uniform succession of days and nights did experience some interruption; the nature and effect of which will best be estimated by considering what would have been the case if it had never happened.

Between a certain νυχθήμερον,

B. C. 1560, inclusive, and the same νυχθήμερον in the time of Joshua, B. C. 1520, exclusive, there would be forty natural years; or 14,609 days and nights, sixteen hours, which I shall consider equivalent to another day and night and consequently, 2087 weeks, and one day of another. Let this vuxenμepov, for argument's sake, be assumed as April 1, which B. C. 1560 fell upon Thursday; and therefore B. C. 1520 would fall upon Friday. In this case, the next vvxonμepov, April 2, ought to have fallen on Saturday; and if the succession of νυχθήμερα went on as before, it would fall upon Saturday.

But let it be further supposed, for argument's sake also, that the miracle in the time of Joshua was wrought upon Friday; and that upon Friday, April 1. The effect of this miracle was that one day as such was prolonged to the length of two; that is, a day of twelve hours was made a day of twenty-four b without affecting the day of the month, or the day of the week; (for April I did not thereby become April 2, nor Friday become Saturday;) but only the absolute length of one individual νυχθήμερον, compared with what the length of every νυχθήμερον was before, and what it continued to be afterwards. The actual April I was Friday, and the actual April 2 was Satur

day; but the actual length of that νυχθήμερον, of which this April 1 was a part, was twelve hours greater than usual.

If then a stranger to this effect were calculating the succession of days and nights from a certain date, before the time of this anomaly, up to a certain date after it; and calculating it upon the supposition that they had always gone on alike, and had always been of uniform length; it is manifest that he would arrive at a conclusion which would be true in theory, but false in fact; viz. that a given vuxenμepov of calculated time began twelve hours later than the same portion of actual time did. He would suppose, for instance, that the vuxenμepov expressed by April 1, B. C. 1520, was an ordinary vuxenμepov of twenty-four hours; whereas it was an extraordinary one of thirty-six and that the next vxonμepov, expressed by April 2, began as usual at the expiration of twenty-four hours of actual time; whereas it did not begin until the expiration of thirtysix. Twelve hours of the calculated second of April were merged in the actual first; and instead of coinciding with Saturday, actually made a part of Friday. But one who was ignorant of this anomaly would suppose they made part of the Saturday, and he would compute them accordingly; that is, his calculated April 2 would be supposed to begin twelve hours later than the νυχθήμερον which it expressed. His calculated April 2 would be reckoned to belong wholly to

Saturday, whereas in reality twelve hours of it were merged in the Friday.

If the effect which ensued in the time of Joshua was repeated in the time of Hezekiah, then another twelve hours of time, which should belong to the calculated νυχθήμερον, would be merged in the actual vyenuepov immediately before it; and both these anomalies together would produce this effect: that reckoning from a certain date before the time of Joshua to a certain date after the time of Hezekiah, and ignorant of each of these miracles, I should suppose a certain calculated νυχθήμερον (we will suppose the third of April) to have been wholly coincident with a certain day of the week, (we will assume the Sunday,) when in fact it was wholly merged in the day before it. That is to say, ever after the miracle in the time of Hezekiah, the actual place of a given vuxenμepov which I might calculate to be Sunday, would be truly the Saturday.

On this principle, April 3, B. C. 4, the place of which was found by calculation to be Sunday, would actually be Saturday; that is to say, the first vvxonμepov of the 208, 710th week, from the Creation, B. C. 4004, which I calculated to begin at sunset on the Sunday, B. C. 4, did actually begin at sunset on the Saturday, B. C. 4 and if I must call that vuxðńμepov April 3, then April 3, which I supposed to be Sunday, was in reality Saturday. Now it makes no difference whether