England from the Revd. John Quicke, M.A., and particularly mentions his Icones Sacrae, then and now in MS. But therein is no allusion to Larkham unless "the vaine janglings about unprofitable questions in ye days of Liberty," which he says occurred in Tavistock after the departure of Rev. George Hughes, are to be referred to Larkham. Quicke died 1706, before the date of the second volume (1727), which contains the matter quoted. He gives no other authorities, so it is impossible now to trace the truth of this story as to Larkham's burial. If it is true that he was buried in the Church, then even after death he was the cause of discord among the people of Tavistock.

I feel that in spite of its length, my paper is a collection of materials, rather than a finished life of Thomas Larkham. But at least I have collected fresh and original matter, and the future historian will find it easier to paint a life-like portrait of the man, than if he had only the previous collected accounts to rely upon. For these, like the later portraits of Queen Elizabeth, paint him " without shadows" and present but a lifeless semblance of the man in his habit as he lived.

ADDITIONAL DISCOVERIES AT THE CASTLE,

EXETER.

BY SIR J. B. PHEAR, M.A., F.G.S.

(Read at Plymouth, July, 1892.)

In a note to the paper on "Recent Discoveries at the Castle, Exeter," which was printed in last year's Transactions of the Association, it was stated that while the sheets were passing through the press, namely, on the 9th October, 1891, about 2 p.m., a good deal of the new work of repair which had been done to the Castle wall, and which was described in the paper, together with the whole of the old inner face of the wall which had been left remaining up to the gateway, suddenly fell without the slightest warning. It is the object of the present paper to continue from this point the narrative of the measures taken to preserve, so far as practicable, this venerable relic of Devonshire feudalism.

It should, however, be mentioned, that before this catastrophe occurred, and while the inner face of the Castle wall was being rebuilt (as explained in the paper), it had been determined, in consequence of the discovery of the human skeletons close under the floors of the Castle-keeper's lodge, coupled with the offensiveness of the soil, and the fact that the house was by reason of its situation incurably damp and well-nigh sunless, that it should be taken down, and a new lodge built on a site somewhat higher up the Castle yard.

The old building was accordingly demolished, and with it doubtless was lost a certain amount of support, which must have been afforded to the inner face of the Castle wall under repair by the walls and roof of a portion of the lodge that abutted against it.

This removal of the lodge disclosed a somewhat worse condition of the wall thus exposed to view than had been anticipated; but at the same time an old buttress was brought to light, on the east side of the Castle gateway, built of

conglomerate, with a weathering, and having on its west face, at some distance from the ground, a portion of a niche, or walled-up arch; and no doubt arose as to the stability of the wall, because the actual character of its internal structure, as afterwards discovered, was not suspected.

About midday, however, on the 9th October, 1891, as mentioned above, a large portion of the old inner face of the wall against which the lodge had stood, reaching in length as far as the Castle gateway, suddenly fell into the Castle yard, carrying with it, of course, such new repairs as had been done. to its upper part. The old buttress did not actually fall to the ground with the rest, but slid bodily in towards the Castle yard, and assumed so dangerous a position of inclination that it had to be removed.

By this fall of the inner face of the Castle wall a longitudinal vertical section of the wall was in effect made, which for the first time disclosed the nature of its construction; and it was at once apparent that from the gateway to at least the middle point of the south front of the Castle the wall was built on the top of an earthen mound or rampart, which had probably been formed of the material thrown up in making the ditch. For a certain distance from the gateway this earthen rampart had been cut down vertically on the inner side to the level of the ground of the Castle yard (a depth of some 14 feet) in order to make room for the site of the keeper's lodge, and had been faced with a retaining wall, which thus constituted the foundation at this point of the inner face of the Castle wall. It was this that gave way after the lodge abutting upon it had been removed, and so brought down all the inner face affected by it. A few weeks afterwards (in November) a portion of the corresponding outer face of the wall also fell, notwithstanding the utmost endeavours made to keep it up by a complete system of shoring. The loose earth on which the standing wall rested slipped away from under it, on the inside, like sand, with the result that this portion of the wall settled downwards, and fell over outwards.

The photographs laid on the table depict the appearance of the wall after the inner casing had fallen, and while the outer portion left standing was shored up; and also show the later state of the case, when the outer portion too had given way. The cross-section of the wall exposed at this stage is very instructive as to its internal structure.

Mr. Harbottle, the County Surveyor, under whose care and superintendence the repairs and alterations rendered necessary by the disaster have been effected, in a memorandum

as to the work done (which he has kindly prepared for me), gives the following description of the old wall:

"The fallen portion of the wall had been built from the level of the foundations of the old lodge, or about the level of the ground on the Castle yard side.

"This wall, for a height of about 14 feet, was built against what is evidently made ground, or ground thrown up, consisting of rubble stones and soil of a very loose description. Above this level (that is, on the top of the mound or rampart) the wall was much thicker, being in section about 8 feet, and constructed with two outer casings, and filled in between the casings with loose inferior packing, consisting of small stones and poor mortar. There is no indication of any proper tie or bond between the two casings, and this fact accounts for the outer or southerly portion being left standing, while the inner or northerly casing separated from it and fell.

"The walling was built with various sizes of stone, principally of two descriptions. One is of a coarse red conglomerate rock, similar to what is obtained from Heavitree quarries, and the other a volcanic trap-stone, probably quarried immediately near the site of the wall, locally known, I believe, as Northernhay stone. The mortar was of the poorest description, with no binding properties left in it.

"The outer facing of the major part of the fallen wall was composed of uncoursed random rubble of volcanic trap, which joined and overlapped at its east end a facing of dressed rangework of the red conglomerate (referred to above) of large stones. The inner or Castle yard side facing was similar to the last description."

With regard to the work of reparation, Mr. Harbottle states that when, in spite of the shoring-up, the outer casing of the wall fell outwards,

"The unsafe walling was then taken down, and rebuilt as nearly as possible with the old facing upon a good concrete platform, and solidly built and tied together, the new stone being obtained from the Heavitree quarries.

"Owing to the nature of the foundation, it was not deemed well to restore the wall to its old width and height, but it was finished with weathered coping, to match that on the existing walling."

The new site selected for the Castle-keeper's lodge is at a short distance towards the north-east from where the old lodge stood, and partially covers the site of the old collegiate chapel, which was taken down in 1792; and in levelling the ground, and digging the foundations for the new lodge, indications of the old chapel floor were seen (as shown by

shading in the accompanying plan), and a good many human skeletons, as well as the lower jaw of a horse, a boar's tusk, two shoulder-pieces of armour, a gold ring, and several silver coins, with one copper coin, were dug up. The situations in which most of these were found are marked on the plan. The chapel floor was about six inches below the level of the castle yard, and two feet below the level of the floor of the new lodge. The skeletons were about nine inches below the floor line of the chapel.

Our retiring President, Mr. R. N. Worth, has been so kind as to examine those of the coins which have come into my possession, and which I was able to send him a few weeks ago for inspection. He pronounces them to be: Four Roman coins, one black money, one English halfpenny of William and Mary, one Irish halfpenny of George II., and one halffarthing of Charles I. Mr. Worth writes me:

"One of the Roman coins bothered me. I believe, however, I am right in ascribing it to Valentinian. There are four Romans in all. Two of the others are Constantines; one certainly, and the other possibly, Constantine the Great, but more likely Constantine II. There is difficulty at times in distinguishing. The fourth is in very bad condition, but I can make out that the head has a radiated crown, and it fits very well with some examples I have of Victorinus; probably either that monarch, or the senior Tetricus.

"The partly broken coin is base metal, and of foreign origin; one of those small pieces which commonly passed current for small change in the fifteenth and sixteenth centuries, often called black money.

"Then there is a half-farthing of Charles I., which I fancy is rather uncommon; the William and Mary halfpenny, and (a thing I never hoped for) I have made out the other to be a George II. Irish halfpenny.

"Altogether a very curious and interesting lot. The Constantines are common at Exeter."

All these coins, to which I have been able to add another Constantine which was found in clearing away the ruins of the wall, are now on the President's table for inspection, and will be deposited in the Albert Museum, Exeter.

As links of evidence carrying our chain of history back nearly 1600 years, they establish an antiquity for the timehonoured fortress of Exeter which ought to strengthen our resolve to let no default of ours prevent this venerable monument of the past from bearing its message undisturbed to the generations yet to come.

R