its side, not a dozen paces from the water. There was no doubt whatever about it. It was roughly five feet long, about half as wide and thick, and of a curious reddish-brown-the colour of dried blood.

'Sir,' said the squire who brought the news to the King and his Knights, 'there is here bynethe at the Ryver a grete stone which I saw flete above the water, and therin I sawe styckyng a swerd. The Kynge sayde, I wille see that marveill. Soo all the Knyghtes went with hym. And whanne they came unto the ryver they fonde there a stone fletyng, as hit were of reed marhel (red marble), and therin stack a fair Ryche swerd.' We confess that not a little awe was mingled with delight as we gazed upon the stone, walked round it, touched it! Then suddenly away in the old city a bell tolled, and we recollected that it was Whitsun Eve! That walk home in the twilight was something not easily to be forgotten, and neither supper nor a pipe could bring us back to earth and the twentieth century again. Next morning we were up early, anxious to see if any trace were left of the spot where this marvel had occurred, for it was scarcely possible that the whole adventure was other than a dream. But the spot was soon found, and sure enough there was the stone or peron,1 and we could examine it

1 Fr. pierron.

in the sunshine at our leisure.

How it got

there or whence it came it were impossible to guess; the chalk for miles around contains nothing but flints, and the peron was smooth and polished as a mill-stone.'

That Winchester is not Camelot antiquaries have told us often enough. The city of the Knights may have been in the West Country, in the Marches of Wales, or in Brittany for aught we care; but until they can produce a likelier site and a better peron we shall continue to take Sir Thomas's word for it.

One other point. We have said that the stone lay some few paces from the water. You will notice when you pay a pilgrimage to the stone (it lies at the ford, hard by a church) that the ground about it is almost level with the water, so that when the river is in flood the stone must be almost submerged in other words, it would then hove above the water. It is easy to see from the bank on the other side that the river has changed its course by a few yards, leaving the stone now high and dry. If you dispute this, why then we can only say that the stone, as 'by adventure it swam down the stream,' must have been cast there by the river when in flood. That there is a cleft in the stone whence Galahad withdrew the sword we can neither affirm nor deny; it may have closed up, for with perons of this nature all things are possible, or the stone itself may have got turned

over.1 At all events we shall not be so rash as to cast suspicion upon so historic a relic.

For those materialists who doubt that such an event ever took place, we will propound a theory. That the first twelve books of the 'Morte d'Arthur' were translated from the French by Sir Thomas Malory seems probable. Caxton says as much in his Preface, and the Epilogue to Book XII. reads, 'Here endeth the second book of Syr Tristram that was drawen oute of Frensshe in to Englysshe. But here is no rehersal of the thyrd book. And here foloweth the noble tale of the Sancgreal that called is the hooly vessel.' It has been shown that the stories of the Holy Grail are probably of Welsh origin, and-Sir Thomas is said to have been a Welshman. Is it possible that he was ever at Winchester, that he wandered on Whitsun Eve (as we did) along the Itchen, that he came to and mused over the stone (smooth and polished as a mill-stone), so different from any to be seen hereabout, and that as he wandered back to Camelot he wove the delicious romance about it? At all events, if he were ever there, it is at least possible that the spot was in his mind when adapting the Welsh legends for his book. Mark how well the events which we relate accord

1 That there is a distinct crack on its upper side, you may see from the photograph here reproduced.

2 Sir J. Rhys, 'Studies in the Arthurian Legend,' Oxford, 1891, pp. 300-327.