¹ Only in one passage is there any hint of a doubt. This is in Mil. Ant., p. 134, where he says that a certain discrepancy is “rather unlucky, as seeming to lessen the dependance we were willing to place on the supposed genuineness of these ancient fragments”.

^{page 162 note 1} Thus, after referring to his plate of the Antonine Wall and its stations, he says that “a short description may suffice, since from a plan of this kind, topographically expressed, a much truer notion may be obtained of the isthmus in general, of every essential particular relating to the wall, and of the military reasons by which the Romans were governed in conducting this boundary of their empire, than what, without such assistance, could possibly be conveyed in many words” (Mil. Ant., pp. 155f.).

^{page 162 note 2} Mr. C. R. Peers and Mr. H. S. Kingsford of the Society of Antiquaries, and Mr. G. F. Hill and Mr. D. T, B. Wood of the British Museum, have helped me not only in this but in many other ways.

^{page 163 note 1} Now preserved in the General Register House, Edinburgh, where I have had an opportunity of consulting them.

^{page 163 note 2} Proc. Soc. Ant. Scot., i, p. 148.Google Scholar

^{page 163 note 3} Addison, W. Innes, Matriculation Albums of the Univ. of Glasgow, p. 29, no. 991.Google Scholar

^{page 163 note 4} Op. cit., vol. lx, p. 670.Google Scholar

^{page 163 note 5} Attention was first drawn to the mistake by Mr. D. R. Rankin, in Proc. Soc. Ant. Scot., ix, p. 564Google Scholar. It may be of interest to supplement Mr. Rankin's correction by indicating the source of the error. The writer in the Gentleman's Magazine was merely copying, in a singularly unintelligent way, a notice which had appeared in Roy's own lifetime in Gough's British Topography (1780), Roy being then a colonel in the army and a captain of engineers. The passage (op. cit., ii, p. 585), which is quoted in full infra, p. 203, shows that Gough was responsible for the mistake in the date and for the undue glorification of the map, as well as for connecting Roy with the artillery.

^{page 164 note 1} p. 64. In the edition of 1888 it is, of course, vol. iii.

^{page 164 note 2} On p. 61 Chalmers had quoted the records of the Privy Council to show that on the date mentioned the Board of Ordnance had represented to the king “the great difficulty of getting proper persons to act as engineers; that the whole establishment of engineers consisted only of 29, of whom 4 were appointed to carry on the works in Scotland”. The king immediately approved of an addition to the establishment of six sub-engineers and ten ‘practitioners’ or probationers.

^{page 165 note 1} The extract quoted above will be found on pp. 385 f. (vol. lxxv).

^{page 165 note 2} Caledonia, ii (iii), p. 62.Google Scholar

^{page 166 note 1} The other two were Hugh Debbieg and George Morrison.

^{page 166 note 2} H. Morse Stephens in Did. Nat. Biogr., xvi, p. 183.Google Scholar

^{page 166 note 3} Through an obvious confusion,Porter, 's History of the Corps of the Royal Engineers, vol. ii (London, 1889)Google Scholar, speaks of Roy as having been Watson's nephew (p. 229).

^{page 166 note 4} It is the large-scale protractions that are lacking. The reduced maps are more complete.

^{page 166 note 5} Vol. lxxv, pp. 386 f.

^{page 167 note 1} p. iv.

^{page 167 note 2} Mil. Ant, p. 197.

^{page 167 note 3} The illustration is reproduced from the British Museum MS., and presents some marked differences from the plate as published: see infra, p. 228.

^{page 167 note 4} Thus Mil. Ant., p. 116, says that “In surveying the line of the border, in 1752, the intrenchments at Chew-green, on the head of Coquet, those at Woden Law, and likewise the tract of the Watling-street, between them, had been taken notice of in the usual way”. On the other hand, in a foot-note on p. 117 Roy expressly disclaims responsibility for the plan of Chew Green as “not having been taken by the author himself”, and goes on to hint a doubt as to its accuracy. Evidently he had no great confidence in the archaeological competence of those of his colleagues who were concerned; speaking of their work on p. 116 he says frankly that “antiquities of this kind were not very particularly attended to”.

^{page 168 note 1} Mil. Ant., p. 116.

^{page 168 note 2} The reference numbers are xlix, 54. 2 (Birrens) and 3 (Birrenswark); see Catalogue of the Manuscript Maps, etc., in the British Museum, (1844) ii; p. 343. The latter agrees exactly with the plan on plate xvi of Mil. Ant. The former is on a slightly smaller scale than the plan on plate xxiv, and there are a few unimportant differences in the representation of the interior; otherwise the correspondence is complete, It is interesting to note that the title of the original sketch employs the spelling ‘Burnswork’, which is undoubtedly the correct form. The variant ‘Birrenswork’ seems to be a later invention of Roy's own, suggested apparently by the analogy of Birrens, and from this came the now generally used ‘Birrenswark’.

^{page 169 note 1} Dr. Christison says that it “is full of errors, and indeed is little better than that of Gordon, which it resembles in making the lines of fortification symmetrical on’ all sides and in misplacing the Portae Principales” (Proc. Soc. Ant. Scot., vol. xxxv, p. 157). It must at the same time be admitted that the language used in Mil. Ant. (p. 122) seems to suggest that Roy had actually visited the spot himself at some time, whether he was the author of the plan or not.

^{page 169 note 2} An interesting account of Melville's life, written by his secretary, was recently published in the Scottish Historical Review (vol. xiv, pp. 116-146). The episode with which we are immediately concerned is, however, only briefly mentioned there. Our information regarding it is drawn from other and even more authentic sources, to be mentioned presently.

^{page 169 note 3} Melville's visit to Edinburgh took place in the autumn. Matthew Clerk (as the manuscript Army List for 1752 in the Public Record Office shows) had received his commission as ensign in the King's Own Borderers in August 1751. He became a lieutenant four years later, and was subsequently transferred to the Engineers. He fell in the unsuccessful attack on Ticonderoga in 1758.

^{page 169 note 4} John Clerk of Eldin, and his younger brother, the ensign.

^{page 170 note 1} The friendship with Lord Panmure, like that with the Penicuik household, appears to have been originally a regimental one. Lord Panmure was colonel of the Twenty-fifth Foot when Melville was a subaltern. He had relinquished the command in 1752.

^{page 170 note 2} Mil. Ant., pp. v f.

^{page 170 note 3} Vol. iii, pp. 414* ff. The reference to the ed. of 1806 is vol. iv, pp. 158 ff.

^{page 171 note 1} Proc. Soc. Ant. Scot, vii, p. 30, where the whole letter is printed.Google Scholar

^{page 171 note 2} This was, of course, the survey.

^{page 171 note 3} Mil. Ant., pp. v f.

^{page 171 note 4} We are told (Mil. Ant., p. 205) that they belonged to a small number of plans or sketches of “very curious British posts” made while he was “searching for, or observing” Roman camps; and they lie only 4 or 5 miles NE. of Keithick.

^{page 172 note 1} They are in the King's Library, the reference numbers being 1. 79, 2 a, 2 b and 3, and 1. 83, 3 (Catalogue of the Manuscript Maps, etc., in the British Museum, ii, pp. 359 f.Google Scholar). They do not show Latin names like Lindum and Hierna. These are derived from Roy's subsequent study, especially of the De Situ Britanniae.

^{page 172 note 2} Mil. Ant., pp. 134 f.Google Scholar

^{page 172 note 3} See infra, p. 213.Google Scholar

^{page 173 note 1} Mil. Ant., p. 155.Google Scholar

^{page 173 note 2} It is perhaps worth noting that Roy's survey, unlike those of Gordon and Horsley, was made from east to west (Mil. Ant., p. 157). This was, as it happens, the direction in which its original designer had worked (Roman Wall in Scotland, pp. 308 f.).

^{page 173 note 3} See, for instance, the last sentence of the passage quoted on p. 171 supra.

^{page 173 note 4} Cf. Mil. Ant., pp. vii ff.Google Scholar

^{page 173 note 5} See infra, p. 208. According to Gough, British Topography, ii, p. 586, it was in the Ordnance Office in 1780. But there is no reason to believe that this statement is any more accurate than the rest of the paragraph in which it occurs: see supra, p. 163 foot-note⁵, and infra, p. 203.

^{page 174 note 1} Roy himself, in the passage quoted supra, p. 165, says that he acted as “assistant Quarter-Master”. Porter in his History of the Corps of the Royal Engineers, ii, p. 229Google Scholar, goes a step farther and makes him “Assistant Quarter-Master-General”. So, too, Portlock in his Memoir of the Life of Major-General Colby (1869), p. 16, where he is spoken of as “General Roy, R.A.”, an echo of the blunder which we have traced back to Gough.

^{page 174 note 2} Vol. ix, pp. 563 f.

^{page 175 note 1} I owe this reference to Mr. John A. Inglis of Auchendinny.

^{page 175 note 2} Col. Vetch in the Diet. Nat. Biogr. (xlix, p. 372) says: “On 23 Dec. 1755 Roy, who had already received a commission in the 4th King's Own Foot, was made a practitioner-engineer.” So far as this statement agrees with what has been said above, it is doubtless drawn from the ‘Record’ of the Royal Engineers, which Col. Vetch cites as one of his sources. The alleged connexion with the Fourth Foot is harder to explain. That the allegation had no foundation in fact seems certain. The manuscript Army List for 1752, which was kept up to date for some years afterwards, and which is now in the Record Office, includes two battalions of the Fourth King's Foot, but Roy's name does not occur in either list of officers. It may perhaps be conjectured that Col. Vetch found the shortness of the interval between Roy's appointment as a practitioner and his commission as a lieutenant inexplicable except on the theory that he had previously been an ensign. In those days the seconding of officers from line-regiments was a recognized means of securing an adequate supply of practitioner-engineers; but, if that had happened in Roy's case, he would undoubtedly have been gazetted to Napier's regiment as ‘Ensign William Roy’, not ‘Engineer’. If this hypothesis as to the origin of the mistake be accepted, it is easy to imagine a confusion between the number of the regiment and the date of the commission.

^{page 177 note 1} Philosophical Transactions, lxxv, p. 387.Google Scholar

^{page 177 note 2} The words are those of a Royal Warrant (War Office Records, Class 55, No. 365, p. 14), dated 31st July 1765, and addressed to the Master of the Ordnance, giving directions for the payment to Roy of an allowance of twenty shillings per day in respect of these duties. Mr. S. C. Ratcliff, of the Public Record Office, who has kindly made search for me, writes that the payment continued to be made quarterly until 31st March 1790 (W. O., 52/37, p. 178), that is, until the last quarter-day before Roy's death. This allowance does not seem to have covered the special missions which he more than once discharged abroad. Thus, W. O. 55/365, pp. 118 f. shows that he received £3 a day for 119 days (26th Oct. 1765-2151 Feb. 1766) on extraordinary service at Dunkirk.

^{page 177 note 3} It is dated 1766. See Notes and Queries, 2nd Series, vii, p. 358. Soon afterwards he was despatched to Gibraltar to report upon the defences.

^{page 177 note 4} Mil. Ant., p. vii.Google Scholar

^{page 177 note 5} The parish register of Carluke has the entry ‘Mortcloth to John Roy’ against the date 25th December 1748, and the corresponding entry ‘Best Mortcloth to Mrs. Roy in Lanark’ against the date 6th July 1777. There can hardly be a doubt that the latter refers to Roy's mother, who would naturally be buried beside her husband. If so, it shows that the widow was settled in Lanark at the time of her death. And the probabilities are all in favour of the removal having taken place early in 1749. John Roy's will, which was executed a week or two before his death, and a copy of which is in the General Register House at Edinburgh, discloses a degree of financial embarassment that must have rendered the breaking up of the old home inevitable.

^{page 178 note 1} Mil. Ant., p. xi.Google Scholar

^{page 178 note 2} Ibid., pp. vii f.

^{page 178 note 3} Ibid., p. vi.

^{page 178 note 4} See supra, p. 171.

^{page 179 note 1} Scottish Historical Review, xiv (1917), p. 121Google Scholar.

^{page 179 note 2} See Mil. Ant., plate vi.

^{page 179 note 3} Mil. Ant, p. viii.

^{page 180 note 1} See Mil. Ant., p. 103, for Roy's view as to the line of the Roman road. That Sir John Clerl was ultimately responsible for the theory may be inferred from a passage in Sir John Sinclair’: Statistical Account of Scotland (vol. x, p. 287), where the writer, after describing certain antiquitie: found at or near Mavisbank, proceeds: "These circumstances led Sir John Clerk, who was we! acquainted with the antiquities of this country, to suppose that this must have been a Roman station And, accordingly, the late General Roy has pointed it out in his maps as the place where the Roman passed the North Esk, in their way from the South to Cramond.”

^{page 180 note 2} Mil. Ant., p. viii.Google Scholar

^{page 181 note 1} Mil. Ant., p. 61.Google Scholar

^{page 181 note 2} Ibid., l. c

^{page 181 note 3} Ibid., p. 115.

^{page 182 note 1} 33. George II. c. 56. There is no copy of this Act in any of the Edinburgh legal libraries, and the title in the printed list does not make it immediately obvious that it is the road past Channelkirk that is intended. Mr. S. C. Ratcliff, of the Public Record Office, has, however, been good enough to send me a transcript of the preamble from the Parliament Roll (33 George II, part 10, m. 1) which puts the matter beyond doubt. It runs: “Whereas the high road leading from Deanburn bridge to the confines of the County of Midlothian through Soutrahill by Channelkirk Greenlaw and Antonhill to the side of the Tweed opposite to Coldstream and from thence to the west end of the village of Cornhill in the County Palatine of Durham is during the winter season or wet weather almost impassable &c. &c.”

^{page 182 note 2} Caledonia, vol. ii (iii), p. 313, and foot-note (c).

^{page 182 note 3} See Curie, , A Roman Frontier Post, pp. 15 ffGoogle Scholar.

^{page 182 note 4} Otherwise he could not possibly have missed “the old course of the way” of whose existence he only learned “on inquiry afterwards made” (Mil. Ant., p. 116); and he would certainly have taken occasion to carry out the “more accurate observation” of Chew Green, which he felt to be desirable (Mil. Ant., p. 117, foot-note).

^{page 183 note 1} Mil. Ant., p. 115.Google Scholar

^{page 183 note 2} Ibid., p. 73.

^{page 183 note 3} p. ix.

^{page 183 note 4} He tells us (Mil. Ant., p. 91) that the idea of making such an attempt was first suggested to him by a remark made by Stukeley in his Commentary on the De Situ Britanniae.

^{page 183 note 5} Mil. Ant., p. 116.

^{page 184 note 1} Mil. Ant., p. 116.

^{page 184 note 2} Ibid., p. viii.

^{page 185 note 1} See infra, p. 224.

^{page 185 note 2} Hist, of Scotland, vol. i, p. 198.Google Scholar

^{page 185 note 3} Cant's ed. of Adamson's Muses’ Threnodie (1774), p. 25. A third series of pits was exposed in April 1774 (Ibid., p. xxi).

^{page 185 note 4} It may have been then that he made inquiry as to whether there were any traces of the supposed “Roman way leading from Glasgow to Paisley” (Mil. Ant., p. 106).

^{page 186 note 1} Roman Wall in Scotland, p. 171.

^{page 186 note 2} Mil. Ant., p. 161.

^{page 186 note 3} Roman Wall in Scotland, p. 327. Roy is vaguer; writing, as we shall see, in 1771 or 1772, he says “some few years ago” (Mil. Ant., I. c).

^{page 186 note 4} Macdonald, James, Tituli Hunteriani, p. 3Google Scholar.

^{page 187 note 1} See supra, p. 181.

^{page 187 note 2} The drawing in the British Museum, it is fair to state, reads “in 1769”, and the preposition accordingly appears in the corresponding ‘List of the Drawings’ there, from which it has been transferred to the printed ‘List of Plates ‘(Mil. Ant., p. 208). The title in Mil. Ant. is, however, an exact reproduction of that in the drawing in the Society's Library, and the drawings in the Society's set are, as a rule, the original sketches (see p. 196, infra). The insertion of’ in ‘in the revised edition may well have been an oversight.

^{page 187 note 3} As we shall see in due course (infra, p. 200 and p. 214), the drawings in Roy's original collection were, in these two cases, entirely different from the published plates.

^{page 188 note 1} See supra, p. 172.Google Scholar

^{page 188 note 2} Mil. Ant., p. 65.Google Scholar

^{page 188 note 3} Ibid., pp. 131 f.

^{page 189 note 1} p. xii.

^{page 189 note 2} Mil. Ant., pp. 160 f.

^{page 189 note 3} They do not seem to have been discovered when Pococke visited the district in 1750. On the other hand, four out of the five are referred to in a letter still extant in manuscript and dated September 25, 1770 (Proc. Soc. Ant. Scot., vol. xxx, pp. 123 f., foot-note). So far as we know, Roy's last sojourn in the neighbourhood was in 1764.

^{page 189 note 4} Anderson's ‘Observations’ bear date ‘January 2d, 1773’ (Mil. Ant, p. 204).

^{page 189 note 5} Mil. Ant, p. 171.

^{page 190 note 1} Mil. Ant., p. 176.

^{page 190 note 2} The reference number is ‘King's MSS. 247, 248’.

^{page 190 note 3} The dates are 4th February and 29th March. See infra, p. 211.

^{page 190 note 4} Some of the corrections are certainly, others possibly, made by the scribe. But there are at least one or two others which are clearly in Roy's own hand, as can be proved by comparison with autograph letters of his, now in the British Museum.

^{page 190 note 5} On the whole, I am disposed to regard the two hands as different

^{page 194 note 1} The original has ‘per Veterum’, which is corrected in the engraving into ‘perveterum’. Roy cannot be acquitted of responsibility for the mistake, since it recurs in the original title of plate ii (see fig. 4, p. 197, infra) and also in the manuscript ‘List of Drawings’, from which latter it even found its way into the printed book (p. 207). The error was natural enough for one who was not a professional scholar, and who was probably indebted to a friend for the titles.

^{page 194 note 2} See supra, p. 172.

^{page 196 note 1} The contrast between pl. XXIX and pl. XXX is instructive.

^{page 196 note 2} A very sketchy reproduction of this design is found in the corresponding drawing in S. A., which is, however, quite possibly a copy, made long after Roy's death: see infra, p. 213).

^{page 199 note 1} See the reproduction in pi XXVII, supra, which should be compared with the Mil. Ant. plate as engraved (plate xlvi).

^{page 200 note 1} The engraver blunderingly calls it ‘Totford’.

^{page 201 note 1} Vol. ii, p. 586. The whole passage is quoted infra, p. 203.

^{page 201 note 2} See supra, p. 173, foot-note⁵.Google Scholar

^{page 202 note 1} See Mil. Ant., p. 117, foot-note.Google Scholar

^{page 202 note 2} Phil. Trans., vol. lxvii, p. 721.Google Scholar

^{page 202 note 3} See supra, p. 178.Google Scholar

^{page 203 note 1} Op. cit., vol. xlix, p. 372.Google Scholar

^{page 203 note 2} Op. cit., vol. ii, p. 561.Google Scholar

^{page 203 note 3} Op. cit., vol. ii, p. 586.Google Scholar

^{page 204 note 1} See supra, p. 163.Google Scholar

^{page 204 note 2} His main change, it will be remembered, was greatly for the worse: he altered “Colonel Roy of the artillery” into “While colonel of artillery, he”, and so started a long train of error.

^{page 205 note 1} Phil. Trans., lxxv, p. 387.Google Scholar

^{page 205 note 2} It was at Buckingham Palace (the old building) that the King's Library (George Ill's) was housed before its transfer to the British Museum.

^{page 205 note 3} It must have been at the same time that there was added to B. M. (p. 58) the last sentence of what has become the first foot-note on p. 51 of the Military Antiquities. It is not found in S. A. (p. 78),

^{page 207 note 1} Elected 12th Jan. 1775.

^{page 208 note 1} Proc. Soc. Ant. Scot., vii, p. 31.Google Scholar

^{page 208 note 2} Curie, , A Roman Frontier Post, p. 140Google Scholar.

^{page 208 note 3} See supra, p. 166.Google Scholar

^{page 208 note 4} Op. at., lxxv, p. 417.Google Scholar

^{page 208 note 5} As already indicated (supra, p. 163), the parish register of Carluke records the baptism of two sisters—Grizel (1723) and Susanna (1728). I have not been able to find any trace of a third. But it would be hazardous to deny the possibility of her existence, and it is worth pointing out that, if a third daughter were born, it would be in accordance with old Scottish custom that she should be given her mother's name of Mary. It may be convenient to. set down here the few facts that are ascertainable as to the career of Reynolds, as kindly collected for me by Mr. John A. Inglis. On 13th August 1784 he was gazetted ensign in the 34th Regiment, which had been in Canada since 1782. He became lieutenant in 1788 and captain in 1791. In the latter year he was ‘disbanded’, remaining on half-pay till 1793, when he secured a captaincy in the first battalion of the ‘Scotch Brigade’. Two years later he gained his majority in the 30th (Cambridgeshire) Foot. He was given the army rank of lieutenant-colonel in January 1799, but he continued to serve on the strength of the 30th Regiment as major until June 1801, when he retired. As, according to the Regimental History, he died in the same year, his retirement was probably due to ill health.

^{page 209 note 1} See supra, pp. 168 and 172.Google Scholar

^{page 209 note 2} See Catalogue of the Manuscript Maps, etc., in the British Museum (1844), ii, p. 332 fGoogle Scholar.

^{page 209 note 3} Caledonia, ii (iii), p. 64Google Scholar. The whole passage is quoted supra, p. 164.

^{page 210 note 1} These were the description of the Antonine Wall (17th, 24th, and 31st March), the account of the Roads (7th and 14th April), Professor Anderson's Appendix (30th June), and the chapter on Agricola's Temporary Camps (7th July).

^{page 211 note 1} Minute of Council of 19th May 1792.

^{page 212 note 1} Proc. Soc. Ant. Scotland, 1916, pp. 328-31 and 351-4.Google Scholar

^{page 212 note 2} See supra, p. 189.Google Scholar

^{page 213 note 1} In the engraving the original title has been amended by omitting ‘Per’ and also by leaving out the periods. In the ‘List of Plates’, however, ‘per’ has been retained (Mil. Ant., p. 207).

^{page 213 note 2} See supra, p. 197Google Scholar. Of course there is always the possibility that the title in S. A. may be Roy's rough draft.

^{page 213 note 3} See supra, p. 198.Google Scholar

^{page 213 note 4} See supra, p. 186Google Scholar.

^{page 214 note 1} The baronetcy dated from 1762. Thomas was created Baron Dundas of Aske, co. York, in 1794, and was the father of Laurence Dundas, first earl of Zetland. I am indebted to Lyon King for this identification.

^{page 214 note 2} He demitted office on 23rd April 1793.

^{page 214 note 3} See supra, p. 187. It is worth noting that Sir Thomas Dundas was in no way related to General Sir David Dundas, Roy's executor, who was the son of a Robert Dundas, a merchant in Edinburgh.

^{page 214 note 4} For details see infra, p. 227.

^{page 214 note 5} p. 19.

^{page 214 note 6} See supra, p. 168.

^{page 214 note 7} p. 105.

^{page 215 note 1} Mil. Ant., p. 102, where a brief foot-note was all that was required. Another foot-note ought to have been appended to the first paragraph on p. 61, pointing out that the addition of Towford had brought up the number of camps in the “irst set” from four to five.

^{page 215 note 2} Supra, p. 200.

^{page 215 note 3} See supra, p. 167, foot-note⁴.

^{page 216 note 1} See supra, p. 194. Some proofs must have been struck off after the new sites had been inserted but before the note was added, thus constituting an intermediate ‘state’. The only example I have seen is in a bound copy of the Military Antiquities now in the Map Department of the British Museum (460. G. 11).

^{page 216 note 2} It is not without interest to note that the minute concludes “and likewise to Geo. Chalmers Esq.”. It must have been on the evidence picked up then that Chalmers founded his charge against the Society of having paid more heed to ‘splendour’ than to ‘accuracy’: see supra, p. 164.

^{page 217 note 1} Scottish Historical Review, vol. xiv, p. 125. He survived until 1809, being at the time of his death, with one exception, the oldest general in the British army.

^{page 217 note 2} See infra, p. 222.

^{page 217 note 3} The drawings were handed over to Basire on 23rd May 1791 (see supra, p. 210). That the engravings were not expected to be ready until May 1794 is plain from the penultimate sentence of the report quoted immediately below.

^{page 218 note 1} See supra, p. 212.

^{page 218 note 2} Under date 16th April 1793.

^{page 219 note 1} 30th March 1792, 25th January 1793, and 16th April 1793.

^{page 219 note 2} I have to thank Mr. James J. MacLehose, LL.D., for advising me on this point. He adds that the book “is printed, approximately, on an Imperial paper of about 70 lbs. weight per ream. The price of that paper now, in ledger quality, is, according to the last list I have and which is a few months old, £8 3s. 3d. a ream.” To-day, therefore, the paper would have cost the Society about £1,200.

^{page 220 note 1} The circumstance that these and other copies were issued loose explains why it is always advisable for purchasers of the Military Antiquities to make sure that the set of plates is complete.