answer to the delusion of Swift, and they must have been more than a very few, or they would not have been sufficient "to save appearances." Setting aside the representations of the poet, upon which no dependence can be placed, except when he bears witness against himself, there is nothing to oppose, and much to confirm the idea that they were the identical "few" which were published in the quarto of 1741. When Swift first collected the letters in May, 1737, he mentioned that they were not much above sixty, and in July, when they had been sent away, and he described his past act in the language of intention, he said they were not above twenty-five. His account in July, when the correspondence was no longer under his eye, and when his failing memory made him forget the departure of Lord Orrery, is far less reliable than his account in May when he was fresh from the task of sorting the letters. A smaller number than he specified appeared in the quarto which, exclusive of the answers of the Dean, contains only forty. This upon an average does not amount to two a year, and the poet, when he had an end to serve, would not have scrupled to call even sixty "few" in comparison with the many that had been written. Swift imagined that the missing letters might have been lost on some of those occasions when he had been compelled to entrust his papers to friends, and Pope may honestly have believed that they were detained by designing persons; but they were never published, while those which were printed have a chasm of seven years, from June, 1716, to August, 1723, or only one year more than the Dean detected in the series he got ready to despatch to Twickenham. The new correspondence, like Pope's previous volumes, was merely a selection, and there is but a single letter of the poet to Swift in 1714, none whatever in 1715, and again but a single letter in 1716. The suppression of the letters in 1717, as in 1715, or even a slip of memory or a slip of the pen with the Dean, both of which had become a frequent occurrence, will account for the slight discrepancy between the chasm in the printed volume, and the chasm which Swift announced. The letter of August, 1723, is the joint letter of Pope and Bolingbroke, which was sent corrected to Curll, and this is followed by a second gap from August, 1723, to September 14, 1725. The extensive hiatus in the correspondence of which Pope was forewarned by Swift, must in all reason be supposed to be the chief deficiency of which Pope complained, though in language coloured to suit his purpose; and when a similar blank exists in the quarto, there is a strong presumption that the letters which he acknowledged had been sent to save appearances, were the same letters of which the book was composed. A kindred circumstance supports the conclusion. The last letter of Pope in the quarto is dated March 23, 1737, which falls in with the fact that the collection was gathered