will that forthwith it be disafforested saving the common of herbage and of other things in the same Forest to them which before were accustomed to have the same."

From Mr. W. Burt's preface to the first edition of Carrington's Dartmoor we learn that other perambulations were made in 1301, 1377, 1557 (a partial one apparently, in regard to the bounds between the Forest and Brent Moor), 1601, 1609, and 1786.

In the table which accompanies this paper I have arranged side by side in columns as many published versions of perambulations of the forest bounds as I have been able to refer to. These various readings are grouped in accordance with resemblances in spelling, &c. For comparison with these there are other references to portions of the forest bounds, the sources of which are indicated in the headings of the lists. Column 3 contains the full text of the version given in Rowe's Dartmoor. Nos. 5, 7, 8, 9, and 10 are partial lists, the names recorded being those which differ in spelling, &c., from those in Rowe's list. These five are taken from a paper in vol. xi. (1879) of our Transactions, by R. Dymond, F.S.A., on "Historical Documents relating to Dartmoor." The places mentioned in the eleventh column, taken from the "Carta Isabellæ de Fortibus," purport to be bounds between the forest and the "maneria de Boclond, Bykelie, and Walkhampton." Similarly the twelfth list gives one account of the boundary between Brent Moor and the adjacent forest. The columns 15 to 23 include a portion of the evidence taken during litigation instituted by certain rectors of Lydford for the recovery of tithes. It may be here mentioned that the list of the "bounds" (in 1240) recorded by Mr. Burt in his preface to Carrington's Dartmoor (1826) is practically identical with that in the second column, taken from the back of the Ancient Map of the Forest, described in vol. v. (1872) by the late C. Spence Bate, F.R.S.; and so it was undesirable to add it to the table. In seven instances Latin names have been translated by Mr. Burt into English; but apart from these alterations, each word is spelt alike in the two records, except in two or three cases, where one letter only in a word may be different: a variation which could easily have arisen during transcription. The correspondence is so close that it would seem as though Mr. Burt must have copied his list from that on the back of the old map.

The map which illustrates this paper is a reduction from the Ordnance map, 1 in. to a mile, and is as nearly as possible half the scale. The present forest boundary is shewn by a

dotted line; and the interrupted lines represent deviations from the present bounds, which I believe, as a result of careful investigation, to represent fairly accurately the ancient limits of the forest.

On previous occasions (Devon Association, 1889 and 1890) I have discussed certain "Bounds" which are included in particular districts of the Moor, but now propose to consider the whole circuit.

COSDON. It is uncertain whether the top, or some part of the foot, of this hill was the original bound at which the jurors began their perambulation. Judging by analogy in the case of other hills, mentioned simply by name without any qualifying term, it must have been the summit.

PARVA HUNDETORRE.--The fact that in 1240 there was a "little Houndtor" shows that there was also a "great Houndtor"; and it is the latter hill which probably retains the name of Houndtor in these days. If this be so, it is evident that it was not in former days a forest bound. This is practically proved by reference to the 1608 document (no. 13), which says that the next bound was " eastward, one mile or more " from Cosdon. The hill now called Houndtor is distant from the summit nearly two miles in a direction S. by W., and about the same distance S.S.E. from the foot of Cosdon at a place where there is a good ford over the Taw at the north end of the Taw Marsh Valley. If this latter point on the river was originally the bound called Cosdon, then a hill crowned by a tumulus west of Raybarrow pool was probably little Houndtor. It is a mile and a quarter from the ford in a S.E. direction. If, however, the top of Cosdon is the true boundary, then Shellstone hill, which is the same distance S.E. from this point, probably represents little Houndtor. This was the view held by the late Mr. Spence Bate.

THURLESTONE, the next bound, is said in the 1608 document to be only of a mile away. This should exclude Watern Tor, which is at present the next bound to Hound Tor, for it is 1 miles distant. On the other hand, from Shellstone Hill to the remains of a so-called sacred circle, near the head of Forder Brook, which I have ventured (Devon Assoc. 1890) to suggest was the original Thurlestone, is only a little more than of a mile. When I fixed upon this circle as the true bound I had not seen Mr. Spence Bate's paper upon the ancient map of the forest. In it he gives very strong reasons indeed for believing that Thurleston was an artificial group of stones; and he named Gidleigh Circle, on

;

Scorhill Down, as being probably Thurleston. His opinion as to the nature of the mark thus tends to confirm my view though I cannot agree in the suggestion that Gidleigh Circle represents it, for this is only 300 yards from what we both. have concluded is the next boundary; and it would be quite unnecessary to have two limiting marks so close. together.

The boundary line of the forest is about 42 miles long; and a simple calculation will show that the average distance between two neighbouring bounds is about 13 miles. Included in the 42 miles are, however, two distances, 5 and 2 miles, where the boundary follows the course of the Wallabrook and the Rattlebrook, respectively. If these be excluded from the calculation, the average is reduced to 1 miles. Actual measurement on the map from one bound to another shows that 50 per cent. of the interspaces lie between the limits of 1 and 1 miles. One distance-King's Oven to Wallabrook Head-is only of a mile; another the ford on the East Ockment to the top of Cosdon-is as much as 2 miles.

The jurors in 1608 did not state their reasons for thinking that Waterdontorr represented the older Thurleston; and, as far as I know, there is no valid reason for concluding that their Waterdontorr is the same as that now called Watern Tor. There are other Water hills, or Waterdons, on the moor; e.g., the hill on the slope of which King's Oven is; and it is quite possible that one of the hills between Cosdon and the North Teign may at that date have been known as Waterdon. It is suggestive that on the Geological Ordnance Map the name Thurleston was inserted on what is now called Kennon Hill, a short distance northwest from the ruins of the circle I believe was known in 1240 as Thurleston.

The 1608 document tells us that the distance from Waterdontorr to the next bound was "near a myle"; and this exactly agrees with the interval between the ruined circle and the point where the Wallabrook joins the North Teign. It will be noticed that in list 18 " Willtorwill" appears for the first time as a bound immediately preceding "Walter torr," i.e., Waterdontorr. In 1826 (see last column) this is called Wildtor well, which is a spot in the upper Wallabrook valley a little east of Wild Tor. It would seem as if this insertion was made with the intention of "backing up" the statement that Watern Tor was the true bound. Unfortunately this view has been accepted, and the consequence is that a very large piece of the north-east region of the moor has been

detached from the ancient forest and included in the neighbouring commons.

WOTESBROKELAKESFOTE, supposed in 1608 to be what was then called Whoodelake, appears as Hood Lake in Westcote (list 14), and subsequently as Woodlake, Hugh Lake, and Rhodelake. In eleven or twelve records it is said to join the Teign, which is identified as the north branch of the stream in list 17; and in list 18 the place of confluence is called Blackstone. I have not succeeded in ascertaining that this last name is still in use in the locality. Anthony Torr, in 1702, when describing the bounds of the east division of the forest, gives as its most northerly point Ting Clapper. The word Clapper is, of course, the moorland term for one of the old uncemented stone bridges; and "Ting" is a common way of spelling and pronouncing the name of the Teign river.

Across this, close to its confluence with the Wallabrook, where the "east quarter" of the forest adjoins the "north quarter," was one of these bridges prior to 1826, when a flood destroyed it. A small specimen still spans the Wallabrook close by. This Teign clapper is the only one which could have been meant, for Teign-head bridge, over which a track leads to Teign-head farm, has never, to my knowledge, been claimed as a boundary even by the most audacious of wouldbe encroachers upon the forest lands. The argument that it could not have existed in 1702, because it must have been built to facilitate access to the lonely farm, is, in the absence of direct proof, worthless; for the bridge over the Blackaven below East Miltor, and that over the Taw at the foot of Steeperton Tor, which are exactly similar in character to Teign-head bridge, lead simply to what were extensive tinstreaming works; and evidences of mining exist also in the North Teign valley above this bridge.

Mr. F. N. Budd, of Batworthy, informs me that, according to the testimony of an old man named Westcott, who died this year considerably over eighty years of age, Teign-head house was built about seventy years ago. Westcott, when a boy, helped to build it, and his father "ripped the stones for it." The evidence that Wotes brokelakesfote is the place of confluence of the Wallabrook with the North Teign seems to be ample.

HEIGHESTONE, Heghstone, Hingestone, or Highstone-the next bound-is said by the 1608 jurors to be near one myle" distant from the last mark. This brings us exactly to the fine "menhir" or "longstone" on Gidleigh Common, half a mile south-west from Kes Tor. It is still an acknowledged

bound of the forest. The fact that the terms Hengheston, Hengston, and Hangeston are used in some records suggests that possibly the cromlech, ruins of which are still to be seen about 200 yards south of the menhir, may have served as the boundary-mark at some period. The massive cover-stone of such megalithic remains might well be described as a suspended or hanging stone. In list 8 the term Hughston is used; in no. 10, Heaston; Fewstone in 17; Hughstone in 18; and in no. 24, Kingstone. Immediately preceding this last name, Stone Tor is stated to be one of the bounds; but this is due to the acceptance of the error, previously mentioned, by which the boundary is supposed to go to Watern Tor, since Stone Tor lies between this last tor and the Gidleigh menhir.

Mr. Spence Bate concluded that Heghstone was a menhir which might have formerly stood between the Teign and the present "Longstone"; but in this I believe he was misled by one of the names under which the next bound occurs. Langestone or Langstone. He seems to have accepted this as undoubtedly the modern "Longstone"; and, if so, a site for Highstone would have to be found between it and the North Teign. The term Highstone, however, equally well describes the character of the existing menhir.

LANGESTONE is in three records the name of the next bound; but it is also called Gotestone and Getestone, Yesceton and Yessetone, "Yeston or Geston or more commonly Hethstone" in 1608; Heath Stone about 1700. Other corrupt forms of these words are Yessetay, Yestelay, and Halfstone.

This place is said by the jurors of 1608 to be "nere one myle" from Highstone. This in a S.E. direction from the menhir brings us to the South Teign, close to its union with a small affluent on its north side between Fernworthy and Metheral. Close by, on both banks of the river, are aboriginal remains: kistvaens on the left, and fine hut circles and a small avenue on the right bank. It is, then, not unlikely that there may also have been here in ancient days a menhir or a group of stones, like Thurlestone or Highstone. On the ancient map this bound is represented like Thurlestone and Highstone as a group or pile of stones, evidently of artificial origin.

On Moll's map of Devonshire (1713), the scale of which is about 10 miles to 1 inch, the only place named on the Moor itself is "Heath Stone"; and it is placed a short distance north of the Moreton and Princetown road, about three miles south-west of Chagford, which agrees pretty