On the whole, the quadrangle is reminiscent of Oxford, though more palatial and less studious. The house is built round a system of these courtyards: first this one, the Green Court, which is the largest and most magnificent; then the second one, or Stone Court, which is not turfed, like the Green Court, but wholly paved, and which has along one side of it a Jacobean colonnade; the third court is the Water Court, and has none of the display of the first two: it is smaller, and quite demure, indeed rather like some old house in Nuremberg, with the latticed window of one of the galleries running the whole length of it, and the friendly unconcern of an immense bay-tree growing against one of its walls. There are four other courts, making seven in all. This number is supposed to correspond to the days in the week; and in pursuance of this conceit there are in the house fifty-two staircases, corresponding to the weeks in the year, and three hundred and sixty-five rooms, corresponding to the days. I cannot truthfully pretend that I have ever verified these counts, and it may be that their accuracy is accepted solely on the strength of the legend; but, if this is so, then it has been a very persistent legend, and I prefer to sympathise with the amusement of the ultimate architect on making the discovery that by a judicious juggling with his additions he could bring courts, stairs, and rooms up to that satisfactory total.

A stone lobby under the oriel window divides the Green Court from the Stone Court. In summer the great oak doors of this second gate-house are left open, and it has sometimes happened that I have found a stag in the banqueting hall, puzzled but still dignified, strayed in from the park since no barrier checked him.

It becomes impossible, after passing through the formality of the two first quadrangles, to follow the ramblings of the house geographically. They are so involved that, after a lifetime of familiarity, I still catch myself pausing to think out the shortest route from one room to another. Four acres of building is no mean matter.

§ iii

Into the very early mediaeval history of the house I do not think that I need enter. It is suggested that a Roman building once occupied the site, and that some foundations which were recently unearthed beneath the larder—evidently one of the oldest portions—once formed part of that construction. The question of dating the existing buildings, however, is quite sufficiently complicated without going back to a building which no longer exists. Nor do I think that the early owners—the Pembrokes, or the Say and Seles—offer the smallest interest; if we knew precisely what parts of the house we owed to them severally it would be another matter, but the mediaeval records are very scanty. It is safe to say, generally speaking, that the north side is the oldest side; it is the most sombre, the most massive, and the most irregular; there are buttresses, battlements, and towers, but no gables and no embellishments—nothing but solid masonry. Up in the north-east corner is the old kitchen, and the old entrances through dark archways at the top of stairways. The passages here, of thick stone, twist oddly, and their ceilings are groined by semi-arches which have become lost and embedded in the alterations to the stone-work. It is a dark, massive, little-visited corner, this nucleus of Knole.

The house, or such portions of it as then existed, was bought from William, Lord Say and Sele, by Bourchier, Archbishop of Canterbury, on June 30, 1456, and it is clear from the numerous bills among the archives at Lambeth Palace that both he and his more notable successor, Cardinal Morton, carried out extensive additions, alterations, and repairs. It is, however, a very difficult task to determine what parts of the building definitely belong to this period, for, what with the additions of the archbishops and the alterations of the later Sackvilles, all is confusion. It would appear, for instance, that upon a foundation of Tudor masonry the Sackvilles constructed the Elizabethan gables which are now so characteristic a feature of the house; but it is less easy to say exactly how much the first Tudor archbishop found there on his arrival of earlier workmanship. A further confusing factor is the great fire which took place in 1623, and is reported to have destroyed a large part of the building—but exactly where, and how much, we cannot say. Nor are the accounts at Lambeth very illuminating:

In divers costs and expenses made this year [1467] for repairing the manor of Knole, carriage for the two cart loads of lathes from Panters to the manor, 14d. For carriage of thirty loads of stone for the new tower, 7d. load = 16/9. Carriage of six loads of timber at 7d. = 3/6. Carriage of one fother of lead from London to Knole, 3/4.

The next year, 1468:

Repairs at Knole. One labourer for 6 days work in the great chamber and the new seler, 2/-. Making of 700 lathes to the new tower, 14d. One labourer 4½ days in the old kitchen, 4d. Item, for 1 j M1 of walle prygge (sic) to the stable and other places, 13d. One cowl to the masonry, 12d.