Chapter VIII: The Results of Caesar’s Invasions of Britain (27)

[2289] _Congrès internat. d’anthr. et d’archéol. préhist._, 1874, i, 579-84.

[2290] _Ib._, p. 579.

[2291] The theory of Mr. Cecil Torr, who holds that the Cassiterides never existed, is virtually identical with that of M. Hildebrand. ‘In the Phoenician language’, he remarks, ‘the word for island is the same as in the Hebrew ... and this word אי is used repeatedly in the Bible for places beyond the sea, as well as ... islands. Most probably the Phoenicians used this word when speaking of the Cassiterides, meaning thereby that these were places beyond the sea [which he identifies with ‘the north-west corner of Spain’]: but the Greeks understood it in another sense, and thus turned these places into islands’ (_Academy_, xlviii, Oct. 26, 1895, pp. 342-3). To the objection that Publius Crassus reached the Cassiterides by sea, Mr. Torr replies that ‘there is nothing to show that his destination was an island’ (_ib._, Nov. 9, p. 390); but Mr. Talfourd Ely (_ib._, Nov. 16, p. 414) pertinently asks whether the word διαβάς, which Strabo uses in describing the voyage of Crassus, can be used of coasting from one point to another on the same shore. Moreover, it is absurd to contend that ‘there is nothing to show’ that Crassus sailed to an island; for Strabo says that the Cassiterides were ten islands; and Mr. Torr is therefore forced, as we have seen (p. 488), to make the incredible assumption that Strabo’s account of the voyage of Crassus is pure fiction.

[2292] _Ency. Brit._, xviii, 1885, p. 806; _Lit. Centralblatt_, 1871, p. 528.

[2293] Pauly’s _Real-Encyclopädie_, iii, part i, 1897, pp. 860, 863.

[2294] _Hist. of Rome_, v, 1894, p. 63 (_Röm. Gesch._, iii, 1889, p. 269).

[2295] ‘The ancient workings for Tin, in the Scilly Islands, are neither deep, nor many, nor large’ (Wm. Borlase, _Observations on the Ant. ... of ... Cornwall_, 1754, p. 30). [In St. Nicholas Island] ‘we found a row of shallow Tin-pits.... These are the only Tin Pits which we saw, or are any where to be seen, as we were informed, in these Islands’ (_ib._, _Observations on the ... Islands of Scilly_, 1756, p. 45). ‘Some Tin might have been found in the low grounds washed down from the Hills.... There may be also Tin-veins in those Cliffs which we did not visit ... as the _Guêl-Hill_ of BREHAR, _Guêl_ Island, the name _Guêl_ (or _Huêl_) in Cornish signifying a Working for Tin’ (_ib._, pp. 73-4). ‘I have been lately informed that, under one of the Cliffs of ANNET, there is a Load, in which there is the appearance of Tin, and that it looks as if it had been work’d’ (_ib._, p. 73, note m). ‘Tin is found in several of the islands ... but there are now no mines in work’ (D. and S. Lysons, _Magna Britannia_, iii, 1814, p. 337). See Addenda, p. 740.

[2296] See H. d’Arbois de Jubainville, _Principaux auteurs de l’ant. à consulter sur l’hist. des Celtes_, &c., p. 42.

[2297] See F. Marx’s article in _Rhein. Mus._, 1, 1895, pp. 321-47.

[2298] _Ora maritima_, 90-8.--

Et prominentis his iugi surgit caput,
(Oestrymnin istud dixit aevum antiquius,)
Molesque celsa saxei fastigii
Tota in tepentem maxime vergit Notum.
Sub huius autem prominentis vertice
Sinus dehiscit incolis Oestrymnicus,
In quo insulae sese exserunt Oestrymnides,
Laxe iacentes, et metallo divites
Stanni atque plumbi.

[2299] See K. Müllenhoff, _Deutsche Altertumskunde_, i, 1890, p. 91.

[2300] _Ora maritima_, 108-9.

[2301] _Ib._, 110-6.--

Haec inter undas multa caespitum iacet,
Eamque late gens Hibernorum colit.
Propinqua rursus insula Albionum patet.
Tartesiisque in terminos Oestrymnidum
Negotiandi mos erat: Carthaginis
Etiam coloni, et vulgus, inter Herculis
Agitans columnas, haec adibant aequora, &c.

[2302] _Rhein. Mus._, 1, 1895, p. 335.

[2303] _Eng. Hist. Rev._, xix, 1904, pp. 139-40, n. 5.

[2304] _Ora maritima_, 94.

[2305] _Hist. of Anc. Geogr._, p. 37.

[2306] Similarly Appian (_De rebus Hisp._, 1) says that the voyage from Spain to the British Isles occupied half a day!

[2307] This page was written before I had read the relevant passage in Kiepert’s _Formae orbis antiqui_, quoted on p. 493, _infra_. Dr. H. Berger maintains (_Gesch. der wissenschaftlichen Erdkunde der Griechen_, iv, 1893, pp. 24-5) that Strabo’s error was due to a misunderstanding of statements about islands situated on the route which the ships engaged in the tin trade followed; but I cannot conceive how such a misunderstanding could have been suggested by the narrative of Crassus.

[2308] _Deutsche Altertumskunde_, i, 1890, p. 92 and note *; ii, 1887, p. 317.

[2309] _Lit. Centralblatt_, 1871, pp. 528-9; _Ency. Brit._, xviii, 1885, p. 806.

[2310] See pp. 500-7, _infra_.

[2311] See also _L’Anthr._, x, 1899, p. 401, n. 2.

[2312] _Formae orbis antiqui,--insulae Britannicae_, 1893.

[2313] _The Cassiterides_, pp. 52-3. See also p. 80, n. *, and pp. 107-8, where Smith makes an ingenious but hardly successful attempt to account for the statement of Strabo, repeated by Ptolemy, that the Cassiterides were ten in number.

[2314] See p. 490, n. 1, _supra_.

[2315] See pp. 489-90, _supra_.

[2316] _Deutsche Altertumskunde_, i, 1890, pp. 91-2.

[2317] _L’Anthr._, iii, 1892, pp. 275-6.

[2318] Σαλύβη, ὅθεν ἀργύρου ἐστὶ γενέθλη. Homer, _Il._, ii, 857.

[2319] _L’Anthr._, iii, 1892, pp. 277-80. For unfavourable criticisms of M. Reinach’s view see O. Schrader, _Reallexicon der indogermanischen Altertumskunde_, p. 993, and _Eng. Hist. Rev._, xix, 1904, p. 140.

[2320] _Rhein. Mus._, xxxviii, 1883, p. 164.

[2321] _Hist. of Rome_, v, 1894, p. 63 (_Röm. Gesch._, iii, 1889, p. 269).

[2322] George Smith (_The Cassiterides_, p. 80) remarks that if Crassus was Caesar’s lieutenant, his discovery of the Cassiterides ‘must have taken place after the time of Julius Caesar’. But Smith forgets that this Crassus died in 53 B.C.

[2323] _B. G._, v, 12, § 5.

[2324] _Folk-Lore_, i, 1890, pp. 91-2.

[2325] Strabo, iii, 2, § 11.

[2326] Groskurd (_Strabonis Erdbeschreibung_, i, 1831, p. 249) translates the passage: ‘dass Iberiens nördliche Küsten gegen Keltike leichtere Vorbeifahrt haben, als wenn man dem Ocean entgegeneschiffe.’ C. Müller, however, in his edition of Strabo (p. 953), rejects Groskurd’s attempt to defend the common text, and holds that we should read τὰ προσαρκτικὰ μέρη τῆς Ἰβηρίας εὐπαροδώτερα εἶναι τοῖς πρὸς τὴν Κελτικὴν κατὰ τὸν ὠκεανὸν πλέουσι,, mentally supplying after εὐπαροδώτερα the words τῶν νοτίων, if indeed they were not in Strabo’s manuscript; and he gives good reasons for believing that Pytheas meant to say what I have stated in the text.

[2327] _Deutsche Altertumskunde_, i, 1890, p. 370.

[2328] _B. G._, iii, 27, § 2.

[2329] _Hist. of Anc. Geogr._, pp. 38-9.

[2330] Unless a child born in 1888 could have been called Mr. Gladstone’s contemporary. Strabo was born about 63, and Crassus died in 53 B.C.

[2331] As Sir George Cornewall Lewis pointed out (_Hist. Survey of the Astronomy of the Ancients_, 1862, p. 452), ‘the Romans ... were not likely to attempt voyages beyond the Pillars of Hercules before ... 146 B.C., whereas after that time the Carthaginians had no ships or factories; Gades had been sixty years in the hands of the Romans; and ever since the end of the Second Punic War the Romans had been able to extort the secrets of the Carthaginians.... The story doubtless originated in the known commercial jealousy of the Carthaginians,’ &c.

M. Salomon Reinach (_L’Anthr._, x, 1899, p. 400) holds that the Romans were anxious to ascertain the maritime route to the Cassiterides because it was cheaper than the overland route. But is it certain that a voyage of more than 2,000 miles would have been cheaper than a land journey of 600?

[2332] _B. G._, ii, 34; iii, 7, § 2.

[2333] Cf. H. Berger, _Gesch. der wissenschaftlichen Erdkunde der Griechen_, iii, 1891, pp. 29, 34.

[2334] See pp. 500-7, _infra_.

[2335] H. Berger (_op. cit._, p. 29) affirms that, according to Strabo (iii, 5, § 11), Crassus saw ‘with his own eyes’ the tin-mining actually going on; but Strabo does not say this.

[2336] Berger (_op. cit._, pp. 34-5) points out that Crassus’s description [was it his?], reproduced by Strabo, puts us in mind of that of Diodorus (pp. 499, 506, _infra_), and may have been suggested to Crassus by a perusal of Diodorus’s authority. R. Zimmermann, on the contrary, argues (_Hermes_, xxiii, 1888, pp. 121-3) that the passage in Strabo is based upon Posidonius. Obviously not the part which relates to Crassus.

[2337] Ἴσως γὰρ δή τινες ἐπιζητήσουσι, πῶς ... οὐδὲν ἐπὶ πλεῖον εἰρήκαμεν ... περὶ τῶν Βρεττανικῶν νήσων, καὶ τῆς τοῦ καττιτέρου κατασκευῆς, ἔτι δὲ τῶν ἀργυρείων καὶ χρυσείων τῶν κατὰ τὴν Ἰβηρίαν,, &c. (iii, 57, §§ 3-4).

[2338] See E. H. Bunbury, _Hist. of Anc. Geogr._, i, 1879, p. 12.

[2339] _Hist. of Anc. Geogr._, p. 38.

[2340] _L’Anthr._, x, 1899, p. 401. Cf. H. d’Arbois de Jubainville, _Les premiers habitants de l’Europe_, i, 1889, pp. 45-6.

[2341] _Engl. Hist. Review_, xix, 1904, p. 140, note.

[2342] _Bibl. Hist._, v. 22, § 2.--Κομίζουσιν εἴς τινα νῆσον προκειμένην τῆς Πρεττανικῆς, ὀνομαζομένην δὲ Ἴκτιν· κατὰ γὰρ τὰς ἀμπώτεις ἀναξηραμένου τοῦ μεταξὺ τόπου, ταῖς ἁμάξαις εἰς ταύτην κομίζουσι δαψιλῆ τὸν καττίτερον. Ἐντεῦθεν δ’ οἱ ἔμποροι παρὰ τῶν ἐγχωρίων ὠνοῦνται καὶ διακομίζουσιν εἰς τὴν Γαλατίαν· τὸ δὲ τελευταῖον πεζῇ διὰ τῆς Γαλατίας πορευθέντες ἡμέρας ὡς τριάκοντα κατάγουσιν ἐπὶ τῶν ἵππων τὰ φορτία πρὸς τὴν ἐκβολὴν τοῦ Ῥοδανοῦ ποταμοῦ.

[2343] See K. Müllenhoff, _Deutsche Altertumskunde_, i, 1890, pp. 471-2; H. Berger, _Gesch. der wissenschaftlichen Erdkunde der Griechen_, iii, 1891, pp. 34-5; and Pauly’s _Real-Encyclopädie_, iii, part i, 1897, p. 860. Müllenhoff justly remarks that the account which Diodorus gives in v, 22 of the mode in which the tin trade was conducted must have been derived from an eye-witness; and that of all the ancient writers Pytheas was the only one who saw with his own eyes what went on at Ictis. Professor Ridgeway assumes that Diodorus’s account of Ictis was borrowed from Posidonius; but the descriptions which Elton (_Origins of Eng. Hist._, 1890, pp. 30-1, 34-5, 92) and Professor Rhys (_Celtic Britain_, 1904, pp. 45-6) have published of the visit of Posidonius to Britain are purely imaginary; for there is absolutely no evidence that he ever crossed the Channel. Elton refers to a passage in the _Solutiones_ of Priscian of Lydia, a writer of the sixth century (_quaest._ vi, p. 571 of F. Dübner’s edition), which proves nothing about Posidonius. See J. Bake, _Posidonii Rhodii reliquiae doctrinae_, 1810; _Fragm. hist. Graec._, ed. C. Müller, iii, 1849, pp. 245-96; R. Scheppig, _De Posidonio_, 1869, p. 7; _Rev. celt._, vii, 1886, p. 378; and M. Dubois, _Examen de la géogr. de Strabon_, 1891, p. 327.

[2344] Professor Rhys (_Celtic Britain_, 1904, p. 45) says that, according to Diodorus, the tin was brought ‘to the outlet of the Rhone, that is to say, to the meeting of the Rhone and the Saone’, &c. But πρὸς τὴν ἐκβολὴν τοῦ Ῥοδανοῦ ποταμοῦ can only mean ‘to the mouth of the Rhône’. Ἐκβολή sometimes means the issue of a river from a mountainous country: it cannot mean that part of a river where it is joined by an affluent; and I doubt whether the professor would seriously maintain that ‘the outlet of the Rhône’ is at Lyons.

[2345] v, 38, § 5.--Πολὺς δὲ καὶ ἐκ τῆς Πρεττανικῆς νήσου διακομίζεται πρὸς τὴν κατ’ ἀντικρὺ κειμένην Γαλατίαν, καὶ διὰ τῆς μεσογείου Κελτικῆς ἐφ’ ἵππων ὑπὸ τῶν ἐμπόρων ἄγεται παρά τε τοὺς Μασσαλιώτας καὶ εἰς τὴν ὀνομαζομένην πόλιν Ναρβῶνα.

[2346] _Nat. Hist._, iv, 16(30), § 104.--Timaeus historicus a Britannia introrsum sex dierum navigatione abesse dicit insulam Mictim in qua candidum plumbum proveniat; ad eam Britannos vitilibus navigiis corio circumsutis navigare. E. H. Bunbury (_Hist. of Anc. Geogr._, i, 1879, p. 603, n. 9) remarks that ‘it is impossible to say what sense we are to attach to the word “introrsus”, upon which the interpretation of the whole passage, in a geographical sense, depends’. I shall show presently (p. 505, _infra_) that only one sense which is not nonsense can be attributed to _introrsum_.

Müllenhoff (_Deutsche Altertumskunde_, i, 1890, p. 471) holds that Pliny confused the distance of Ictis from Britain with that of Thule, which, as he says in an earlier passage (_Nat. Hist._, ii, 75 [77], § 187), was ‘six days’ sail northward from Britain’ (_sex dierum navigatione in septentrionem a Britannia_). See p. 505, _infra_.

[2347] The geographical position of Corbilo cannot be fixed. Desjardins (_Géogr. de la Gaule rom._, i, 1876, p. 288) was originally inclined to place it near Beslon in the peninsula of Guérande, because the neighbourhood is ‘rempli de souvenirs celtiques’. Beslon is no more on the Loire than Margate is on the Thames; and if the tin had been landed there, it would have been necessary either to tranship it and carry it across the Loire, or to take the pack-horses by a roundabout route up the valley of that river. Afterwards (_ib._, ii, 1878, pp. 139, 484-5, 485, n. 1) Desjardins changed his mind, and identified Corbilo with St.-Nazaire: ‘cet emplacement’, he remarked, anticipating one of the objections which I have just made against his former view, ‘cet emplacement s’accorde-t-il beaucoup mieux que celui de Beslon avec le texte de Strabon, qui porte cet ancien port sur la Loire, et non sur la mer.’ He relied mainly upon the investigations of an engineer, M. René Kerviler, who, ‘ayant eu l’occasion de faire des travaux d’approfondissement à Saint-Nazaire, y a découvert des substructions qui avaient fait vraisemblablement partie de l’ancien port de _Corbilon_.’ See _Rev. arch._, nouv. sér., xxxiii, 1877, pp. 145-53, 230-9, 342-53. M. Kerviler himself identified the remains with those of the _Brivates portus_ of Ptolemy, _Geogr._, ii, 8, § 1.

[2348] _Geogr._, iv, 2, § 1.--πρότερον δὲ Κορβιλὼν ὑπῆρχεν ἐμπόριον ἐπὶ τούτῳ τῷ ποταμῷ, περὶ ἧς εἴρηκε Πολύβιος, μνησθεὶς τῶν ὑπὸ Πυθέου μυθολογηθέντων, ὅτι Μασσαλιωτῶν μὲν τῶν συμμιξάντων Σκιπίωνι οὐδεὶς εἶχε λέγειν οὐδὲν μνήμης ἄξιον ἐρωτηθεὶς ὑπὸ τοῦ Σκιπίωνος περὶ τῆς Βρεττανικῆς, οὐδὲ τῶν ἐκ Νάρβωνος οὐδὲ τῶν ἐκ Κορβιλῶνος, αἵπερ ἦσαν ἄρισται πόλεις τῶν ταύτῃ.

[2349] Cf. _Folk-lore_, i, 1890, pp. 85-6, and H. F. Tozer, _Hist. of Anc. Geogr._, p. 36.

[2350] Cf. K. Müllenhoff, _Deutsche Altertumskunde_, i, 1890, p. 471, and D. Detlefsen in W. Sieglin’s _Quellen und Forschungen_, &c., Heft 9, p. 77.

[2351] _Origins of Eng. Hist._, 1890, p. 34.

[2352] _Celtic Britain_, 1904, p. 46.

[2353] Mr. Alfred Tylor blunders even more hopelessly than Elton. ‘The transhipment of tin’, he says (_Archaeologia_, xlviii, 1885, p. 233), ‘was described by ancient writers as taking place at Vectis, six days’ sail from Cornwall.’

[2354] _Folk-Lore_, i, 1890, pp. 95-7.

[2355] _Ib._, pp. 98-101.

[2356] See pp. 250, 359-60, _supra_.

[2357] Mr. Alfred Tylor (_Archaeologia_, xlviii, 1885, p. 233) argues, in favour of the identification of Ictis with the Isle of Wight, that ‘Stans Ore Point is said to be named from Stannum (tin)’; and Elton (_Origins of Eng. Hist._, 1890, p. 230) thinks that ‘the course of the metal-trade may be indicated by the names of places on the coast-road leading eastward from the Exe, as ... Stans Ore Point’. Now, as O. Schrader points out (_Prehist. Ant. of the Aryan Peoples_, 1890, p. 217), _stannum_ probably did not get the meaning of ‘tin’ before the fourth century A.D.; and even if the derivation in question could be established, it would not prove that Ictis was the Isle of Wight. Tin was doubtless conveyed eastward from Cornwall; but not for the supply of the Mediterranean markets.

[2358] _Archaeologia_, xlviii, 1885, p. 236.

[2359] vi, 2, § 6.--ᾬκουν δὲ καὶ Φοίνικες περὶ πᾶσαν μὲν τὴν Σικελίαν ἄκρας τε ἐπὶ τῇ θαλάσσῃ ἀπολαβόντες καὶ τὰ ἐπικείμενα νησίδια ἐμπορίας ἕνεκεν τῆς πρὸς τοὺς Σικελούς.

[2360] B. Jowett, _Thucydides translated into English_, i, 1881, p. 409.

[2361] _Report of ... the Brit. Association_, 1865 (1866), p. 71.

[2362] _Trans. Roy. Geol. Soc. Cornwall_, iii, 1828, pp. 91-4.

[2363] _Principles of Geology_, i, 1875, pp. 546-7.

[2364] The italics are mine. Müllenhoff (_Deutsche Altertumskunde_, i, 1890, pp. 471-2) asserts that ‘Ictis can only be looked for at the promontory of Belerium’ [the Land’s End], and that ‘it is undoubtedly one of the small islands off the Land’s End, which are marked on the Ordnance Map (sheets 32 and 33)’. It must be presumed that Müllenhoff came to this singular conclusion because Pytheas landed at Belerium. But there is no reason to suppose that he landed at the precise spot which we call the Land’s End; and if he did he certainly went on to visit the tin mines. If Müllenhoff had known the Cornish coast, or even studied the map carefully, he would have seen that tin could not have been conveyed in carts down the cliffs opposite the small islands to which he refers, and that, as Dr. Barham says (_Trans. Roy. Geol. Soc. Cornwall_, iii, 1828, p. 91), ‘there is not ... any other island [besides St. Michael’s Mount] on the Cornish, or any neighbouring shores to which carts can pass at low water; there is no other spot, at all answering to the description of Diodorus, which becomes alternately an island and a peninsula with the changes of the tide.’

George Smith (_The Cassiterides_, p. 114) points out that ‘twelve miles to the west of St. Michael’s Mount, and eighteen miles to the east of it, comprehend almost the whole of the ancient tin mining district’. Professor Rhys, on the other hand, states (_Celtic Britain_, 1904, p. 44) that the tin districts ‘in ancient times were chiefly Dartmoor, with the country around Tavistock, and that around St. Austell, including several valleys looking towards the southern coast of Cornwall’; and he adds that ‘in most of the other districts where tin existed it is supposed to have lain too deep to have been worked in early times’. I do not know whether among these ‘other districts’ he includes the one near St. Michael’s Mount; but it is certain that the tin in this district was worked in early times. It was the district of Belerium, where the tin-workers mentioned by Diodorus lived; and he says that there were veins of tin in the hard rock near the surface (αὕτη δὲ πετρώδης οὖσα διαφυὰς ἔχει γεώδεις, ἐν αἷς τὸν πόρον κατεργαζόμενοι καὶ τήξαντες καθαίρουσιν [v, 22, § 2]. Cf. Strabo, iii, 5, § 11, and _Ency. Brit._, 9th ed., vi, 425). Mr. P. W. Flower (_Hist. of the Trade in Tin_, 1880, p. 26) tells us that from pre-Roman days ‘Cornish men have been sinking deeper and deeper in their search for cheaper metal’; while Prof. Haverfield (_Proc. Soc. Ant._, 2nd ser., xviii, 1900, p. 122), after remarking that ‘the tin districts of Dartmoor [were] worked largely in the middle ages’, says, ‘The Dartmoor tin is, I believe, far more difficult to work than the Cornish, and this fact may explain the Roman neglect of it.’ See also, for evidence that Cornish tin was won in the Bronze Age, _Archaeologia_, xvi, 1812, p. 137, pl. 10; xlix, 1885, p. 181; and _Archaeol. Journal_, xxxi, 1874, pp. 53, 60. I am astonished to find that M. Salomon Reinach (_L’Anthr._, xvii, 1906, pp. 235-6), noticing a paper the writer of which maintains that no tin was worked in Britain until after the date of Domesday Book, says, ‘Cette manière de voir, bien que contredite par les textes, mérite réflexion.’

[2365] _Archaeologia_, lix, part ii, 1905, pp. 281-8.

[2366] _Principles of Geology_, i, 1875, pp. 543-4.

[2367] _Report of ... the Brit. Association_, 1865 (1866), p. 71.

[2368] _Geol. Mag._, 1879, pp. 74-5.

[2369] _Chips from a German Workshop_, iii, 1870, pp. 330-57. Elton, even in his second edition (_Origins of Eng. Hist._, 1890, p. 37), repeated the obsolete argument alluded to in the text.

[2370] See p. 31, _supra_.

[2371] See p. 222, _supra_.

[2372] iv, 16 (30), §§ 103-4. Prof. Ridgeway (_Proc. Soc. Ant._, 2nd ser., xx, 1904, p. 343) affirms, Prof. Rhys (_Celtic Britain_, 1904, p. 304) apparently denies that _Ictis_ and _Vectis_ were phonetically connected. See Addenda, p. 740.

[2373] See p. 499, n. 5, _supra_.

[2374] _Guide to the Ant. of the Early Iron Age_ (Brit. Museum), p. 85.

[2375] See p. 221, _supra_.

[2376] _Dicuili liber de mensura orbis terrae_, ed. G. Parthey, 1870, pp. 42-4 (7, 11-4). Dicuil was an Irish monk, who wrote A.D. 825.

[2377] _Guide to the Ant. of the Bronze Age_ (Brit. Museum), p. 146.

[2378] Φιλόξενοί τε διαφερόντως εἰσὶ καὶ διὰ τὴν τῶν ξένων ἐμπόρων ἐπιμιξίαν ἐξημερωμένοι τὰς ἀγωγάς.

[2379] This is admitted, or rather maintained, by Prof. Ridgeway.

[2380] See p. 499, nn. 2 and 5, _supra_.

[2381] _Deutsche Altertumskunde_, i, 1890, p. 223.

[2382] _Folk-Lore_, i, 1890, p. 105.

[2383] See p. 501, _supra_.

[2384] _B. G._, iii, 8, § 1. Cf. Strabo, iv, 4, § 1.

[2385] _Geogr._, iv, 5, § 1.

[2386] I cannot see how Mr. Reginald Smith (_Guide to the Ant. of the Early Iron Age_ [Brit. Museum], p. 85) reconciles his theory, that the route in ‘the opening years of the first century B.C.’ passed through Kent with his previous assertion (p. 84) that ‘about 90 B.C.’ it left the British coast at the Isle of Wight.

[2387] As Professor Ridgeway assumes that Posidonius was the authority whom Diodorus followed both in v, 22 and in v, 38, he would be compelled to maintain that in the passage which served as the basis of the former chapter Posidonius was describing only the route which the tin trade followed in the time of Pytheas, in the other that which it followed in his own time. How can the professor prove this?

[2388] _Proc. Soc. Ant._, 2nd ser., xviii, 1900, p. 119.

[2389] _Celtic Britain_, 1904, pp. 47-50.

[2390] _Folk-Lore_, i, 1890, pp. 83-4.

[2391] _Archaeol. Journal_, xlix, 1892, p. 178.

[2392] _Proc. Soc. Ant._, 2nd ser., xviii, 1900, pp. 119-20.

[2393] _Proc. Soc. Ant._, 2nd ser., xviii, 1900, p. 122.

[2394] Numerous Roman inscribed objects of lead have been discovered in Spain (_Corpus Inscr. Lat._, ii, 4964, and Suppl., 6243, 6247-8); but so far as I can ascertain, none of tin.

[2395] _Archaeol. Journal_, xlvii, 1890, p. 232.

[2396] _Proc. Soc. Ant._, 2nd ser., xviii, 1900, pp. 119-20.

[2397] _Archaeol. Journal_, xlvii, 1890, pp. 230-3; xlix, 1892, p. 178; _Corpus Inscr. Lat._, vii, 13.

[2398] _Proc. Soc. Ant._, 2nd ser., xviii, 1900, p. 120.

[2399] _Ib._, p. 122.

[2400] _Ib._, p. 118.

[2401] _Ib._ What puzzles me is how Professor Haverfield reconciles his view that in the third century ‘Cornish tin began to take its place as an article of commerce in Roman Britain’ (_Mélanges Boissier_, 1903, p. 251) with his own suggestion (_ib._, p. 250) that ‘either the tin ores had never been so rich as fancy painted, or the accessible deposits had been worked out [two centuries earlier], or ... Spanish competition had ousted British tin’. Evidently the accessible deposits had not been worked out; and British tin must have had superabundant vitality if it reasserted itself two centuries after it had been ousted.

[2402] _Hist. Survey of the Astronomy of the Ancients_, pp. 451-5.

[2403] _Report of ... the Brit. Association_, 1896, p. 910.

[2404] _Anc. Bronze Implements_, p. 419. Cf. F. J. Haverfield in _Mélanges Boissier_, p. 249, n. 1. Mr. Reginald Smith (_Guide to the Ant. of the Early Iron Age_ [Brit. Museum], p. 137) suggests, with the approval of Mr. C. H. Read, that a bronze statuette, found near Aust-on-Severn, may have been deposited ‘by Phoenician traders to our shores’. Cf. _Proc. Soc. Ant._, 2nd ser., xx, 1904-5, p. 192.

[2405] See G. Smith, _The Cassiterides_, p. 54.

[2406] G. Smith, _The Cassiterides_, pp. 47-9. See also pp. 56-7, and E. H. Bunbury, _Hist. of Anc. Geogr._, i, 12.

[2407] _Géogr. de la Gaule rom._, i, 329.

[2408] See pp. 484-6, _supra_.

[2409] _Tin Mining in Spain_, p. 28.

[2410] _Ora Maritima_, 113-6.--

Tartesiisque in terminos Oestrymnidum
Negotiandi mos erat: Carthaginis
Etiam coloni, et vulgus, inter Herculis
Agitans columnas, haec adibant aequora.

See also H. F. Tozer, _Hist. of Anc. Geogr._, pp. 110-1.

[2411] Prof. Haverfield (_Eng. Hist. Rev._, xix, 1904, p. 746) thinks that ‘the “Periplus” of Avienus cannot safely be attributed to Himilco’; but M. Camille Jullian (_Ann. de la Faculté des lettres de Bordeaux,--Bull. hisp._, v. 1903, p. 109; _Journal des Savants_, nouv. sér., No. 2, 1905, pp. 95-8) supports my view. I am not sure, however, that Prof. Haverfield means to express a doubt whether the Periplus was _ultimately_ based upon Himilco’s report. Cf. _Rhein. Mus._, l, 1895, p. 336.

[2412] _Hist. Survey of the Astronomy of the Ancients_, p. 455.

[2413] E. Hübner, _Monumenta linguae Ibericae_, 1893, p. xxvi; H. d’Arbois de Jubainville, _Principaux auteurs de l’ant. à consulter sur l’hist. des Celtes_, &c., p. 42.

[2414] _Nat. Hist._, ii, 67, § 169.--Hanno Carthaginis potentia florente circumvectus a Gadibus ad finem Arabiae navigationem eam prodidit scripto, sicut ad extera Europae noscenda eodem tempore Himilco. I find that Müllenhoff (_Deutsche Altertumskunde_, i, 1890, pp. 93-5) has anticipated a remark which I was about to make, namely, that the object of Himilco’s voyage was undoubtedly to open up new markets for trade, and not merely to explore. See also Lord Avebury’s _Prehist. Times_, 1900, pp. 57-67, though I think that his argument might have been more valuable if he had taken note of Mr. Borlase’s _Tin Mining in Spain_.

[2415] Mr. W. C. Borlase (_Tin Mining in Spain_, pp. 24-6), remarking that ‘there is an extremely rare form of [the palstave], namely with two loops, and that has been found exclusively in Cornwall and Devon (in the mining districts especially), in Ireland, and in the western and north-western portion of the Iberian Peninsula’, and that ‘bronze celts of this class belong ... to ... 1250 to 1050 B.C.’, concludes that ‘at that period then--the very period to which has been assigned the foundation of Gades--Cornwall and the west coast of Spain were already in communication’. Perhaps; but not necessarily Cornwall and Gades. Similar celts have also been found in France (J. Evans, _Anc. Bronze Implements_, pp. 96-7).

Müllenhoff also argues (_Deutsche Altertumskunde_, i, 1890, pp. 5-8) that the passage in the _Odyssey_ (x, 81-6) which describes the country of the Laestrygones, where the days in summer were very long and the nights very short, would seem to be based upon stories told by Phoenician mariners; but, as I have already remarked (p. 218), if Homer’s lines were founded upon fact, it is more probable that the stories came to him from Scandinavia.

[2416] _Report of ... the Brit. Association_, 1896, p. 910.

[2417] _The Builder_, Aug. 26, 1865, p. 604.

[2418] C. F. Wiberg (_Der Einfluss der klassischen Völker_, &c., 1867, p. 13) thinks that ‘the promontory of Herakles’, or Hartland Point (Ptolemy, _Geogr._, ii, 3, § 2), may owe its name to the Phoenician worship of Hercules; but I do not know that any one except Prof. Boyd Dawkins (_Early Man in Britain_, p. 461) attaches any importance to this suggestion.

[2419] _L’Anthr._, x, 1899, p. 401.

[2420] i, 13, § 5.--οἰκοῦντες γὰρ τὴν πόλιν οἱ Κορίνθιοι ἐπὶ τοῦ ἰσθμοῦ ἀεὶ δή ποτε ἐμπόριον εἶχον, τῶν Ἑλλήνων τὸ πάλαι κατὰ γῆν τὰ πλείω ἢ κατὰ θάλασσαν, τῶν τε ἐντὸς Πελοποννήσου καὶ τῶν ἔξω, διὰ τῆς ἐκείνων παρ’ ἀλλήλους ἐπιμισγόντων, &c.

[2421] B. Jowett, _Thucydides translated into English_, i, 1881, p. 10.

[2422] See p. 126, _supra_.

[2423] See p. 485, n. 5, _supra_.

[2424] _Nat. Hist._, vii, 56 (57), § 194.

[2425] _Fabulae_, ed. M. Schmidt, 1872, CCLXXIV (p. 149).

[2426] _Variarum_ iii, 51 (J. P. Migne, _Patrologiae cursus completus_, lxix, 1848, col. 594).

[2427] See C. Müller’s edition of Diodorus, i, 1842, p. 316 (Reliquiae libri vii, 13). The commencement of the maritime supremacy of the Phoenicians is here dated 58 years after the commencement of that of the Phrygians, and 279 years after the Trojan War.

[2428] _Archaeol. Journal_, xxxix, 1882, p. 18; _Essex Naturalist_, i, 1887, pp. 266-76.

[2429] _Trans. Essex Archaeol. Soc._, N. S., vii, 1900, p. 252.

[2430] See pp. 151, n. 4, 253, 256, _supra_.

[2431] _Essex Naturalist_, i, 252.

[2432] _Trans. Essex Archaeol. Soc._, N. S., vii, 253-4.

[2433] _Nat. Hist._, xvii, 8 (4), § 45. Cf. _Essex Naturalist_, i, 249.

[2434] _Ib._, pp. 249-50.

[2435] _Journ. Brit. Archaeol. Association_, N. S., x, 1904, pp. 98-101.

[2436] _Essex Naturalist_, i, 250.

[2437] _Archaeol. Journal_, xxxix, 1882, p. 19.

[2438] _Journ. Brit. Archaeol. Association_, N. S., x, 1904, p. 98. Cf. _Times_, Sept. 30, 1905, p. 3, cols. 3-4.

[2439] _Essex Naturalist_, i, 250-1.

[2440] _Geol. Mag._, 1898, p. 453.

[2441] _Geol. Mag._, 1898, p. 453.

[2442] Worthington G. Smith, _Man, the Primeval Savage_, pp. 326-7.

[2443] J. A. H. Murray, _New Eng. Dict._, iii, 192-3.

[2444] _Geol. Mag._, 1898, p. 457.

[2445] _Trans. Essex Archaeol. Association_, N. S., vii, 1900, p. 253.

[2446] _Vict. Hist. of ... Essex_, i, 310-1.

[2447] _Excavations in Cranborne Chase_, i, 4.

[2448] A. Joanne, _Dict. géogr. ... de la France_, 1869, p. xli.

[2449] _Géogr. de la Gaule rom._, i, 349 and n. 3, 391.

[2450] In particular C. de Laroière in _Annales du comité flamand de France_, x, 1868-9 (1870), pp. 249-322.

[2451] See _Bull. de l’Acad. Roy. ... de Belgique_, 3^e sér., viii, 1884, pp. 681-9. Desjardins (_Géogr. de la Gaule rom._, i, 391) admits that it is only ‘probable’ that it existed at that time. Cf. R. Blanchard, _La Flamande_, 1906, pp. 134-46.

[2452] _Rev. sc._, 2^e sér., xv, 1878-9 (1879), pp. 90-3. According to M. V. J. Vaillant (_Classis Britannica_, 1888, pp. 66-7), an inscription (_SALINATORES CIVITATIS MORINORVM_), ‘cité par I. Gruter, nous rapporte à une époque où la mer pénétrait librement jusqu’au delà de Saint-Omer et où les marais salants étaient exploités sur les rives de ce large golfe par les Morins et les Ménapiens.’ M. Vaillant gives neither the reference nor the date of the inscription: it was found at Ariminum in Cisalpine Gaul, and makes mention of the emperor Vespasian; and it is reproduced in Gruter’s _Inscr. ant. totius orbis Romani_, ii, 1707, p. MXCVI, 4. Needless to say, it does not prove that the ‘gulf’ existed in Vespasian’s time, but only that there were salt-works in the territory of the Morini.

[2453] See pp. 565-7, 572, 586-7, _infra_.

[2454] _Boulogne-sur-mer et la région boulonnaise_, i, 359-61.

[2455] A. Joanne, _Dict. géog. ... de la France_, p. xlii.

[2456] _Hist. eccl._, i, 25. Cf. Solinus, ed. Th. Mommsen, p. 114.

[2457] _Archaeol. Cant._, xii, 1878, p. 3. See also _Twenty-third Report East Kent Nat. Hist. Soc._, 1881, p. 48.

[2458] _Ant. of Richborough_, &c., 1774, pp. 137-9; _Archaeologia_, i, 1770, pp. 79-83.

[2459] _Hist. of Sandwich_, 1792, p. 865.

[2460] _Archaeol. Journal_, liii, 1896, p. 207.

[2461] _Itin. Ant._, ed. P. Wesseling, 1735, p. 472.

[2462] The reference is incorrect. For ‘30’ read ‘230’.

[2463] _Archaeol. Journal_, liii, 1896, p. 207.

[2464] Sheet 290.

[2465] _Archaeol. Journal_, liii, 1896, pp. 212-3.

[2466] _Archaeol. Cant._, xxiv, 1900, p. 110.

[2467] _Ant. of Richborough, Reculver and Lymne_, 1850, pp. 53-4. See also _Archaeol. Cant._, xiv, 1882, pp. 368-9; xxiv, 1900, p. 108; and _Archaeologia_, li, 1888, p. 465. Beale Poste (_Britannia antiqua_, 1857, p. 282) states that in one of the sand-hills, half a mile north of Sandown Castle, a large number of coins of Victorinus, Probus, Tetricus, ‘and others of the lower empire’ were found in 1839.

[2468] See _Archaeol. Cant._, viii, 1872, pp. 13-4. Boys, quoted by Roach Smith (_Ant. of Richborough_, &c., p. 53) remarks that, ‘in digging to lay the foundation of Richborough sluice, the workmen, after penetrating through what was once the bed of the river that runs close by ... came to a seashore that had been suddenly covered with silt.’

[2469] ‘Just north of the Isle of Richborough’, says Dowker (_Journ. Brit. Archaeol. Association_, xl, 1884, p. 272), ‘is a large artificial excavation in the hill. I gave a description of this when I wrote the account of the ... excavation at the Castrum; and I drew attention to its being a Roman harbour. It is just opposite a farm that goes by the name of “Fleet”.’

[2470] _Archaeol. Journal_, liii, 1896, p. 356.

[2471] _Ib._, xxxiii, 1876, p. 71.

[2472] _Ib._, facing page 64.

[2473] _Archaeol. Cant._, xxiv, 1900, p. 108.

[2474] _Dict. Nat. Biogr._, xlviii, 15.

[2475] _Archaeologia_, xxi, 1827, p. 505.

[2476] _The Cinque Ports_, 1888, p. 229.

[2477] I am glad to find that this remark has been anticipated by Mr. C. R. S. Elvin (_Records of Walmer_, 1890, p. 30).

[2478] _Itinerary_, 2nd ed., vii, 1744, fol. 127 (p. 116). Professor Burrows may perhaps have followed Hasted, who says (_Hist. of Kent_, iv, 1799, p. 163) that ‘_Upper Deal_ was composed of the habitations of a few poor fishermen only, though at a less distance from the sea than at present, owing to the great increase of beach thrown on this shore afterwards’; and in note _e_ he observes that ‘Leland ... seems to confirm this’. Leland, as I show in the text, does no such thing. Hasted goes on to say that ‘Even so late as the year 1624, a house ... on the _west_ side of the _Lower Street_ (the farthest at this time from the sea shore) is described in a deed of that date to abut _ad le sea bank versus orientem_’. Very likely: but the fact does not prove that the west side of Lower Street was an inch nearer the sea in 1624 than it is now; for the breadth of ‘_le sea bank_’ is not stated. Anyhow Deal Castle has not moved since 1624: therefore, if Hasted is right, the sea must then have made a sudden bend landward immediately north of Deal Castle, and formed a bay; which is absurd. The west side of Lower Street is now about 550 feet from the high-water mark of ordinary tides (_Six-Inch Ordnance Survey_, Sheets 58 and 58A).

[2479] The distance from the ‘high-water mark of ordinary tides’ to the nearest point of Upper Deal appears to be about 3,900 feet (_Six-Inch Ordnance Survey_, Sheet 58).

[2480] _Archaeol. Journal_, xxxiii, 1876, p. 71.

[2481] _Ib._, p. 58.

[2482] _Ib._, p. 59.

[2483] _Proc. Geologists’ Association_, ix, 1885-6 (1887), pp. 174-5.

[2484] ‘It is certain,’ wrote Dowker in 1876 (_Archaeol. Journal_, xxxiii, 59), ‘that when the sea swept the Stonar beach, Deal had no existence.’ Even men of science sometimes use the word ‘certain’ a little rashly. At that time Dowker asserted that the Stonar beach ‘must have travelled from the cliff between Dover and Deal’. In 1887 (_Proc. Geologists’ Association_, ix, 174-5) he ‘pointed to the stones of which it is composed as evincing their origin from the cliff at Pegwell.... To imagine it to have travelled from the south, we must,’ he said, ‘have a shore-line cutting far back beyond the Deal beach, of which at present there was no evidence.’

[2485] _Itin. curiosum_, 2nd ed., 1776, pp. 126-7.

[2486] It must be borne in mind that Stukeley wrote before the great increase of shingle in the neighbourhood of Walmer.

[2487] _Ib._; C. R. S. Elvin, _Records of Walmer_, pp. 2-3, 5.

[2488] _Calendar of State Papers, Domestic Series_, of James I, 1611-8, p. 324 (vol. lxxxii, 129), under date 1615; _ib._, Charles I, 1625-6, p. 321 (vol. xxv, 82), under date 1626; _ib._, 1627-8, p. 200 (vol. lxv. 62), under date 1627. In the British Museum is a print, called ‘N.W. View of Deal Castle’, published in 1735, from which it would appear that at that time the castle was as close to the sea as it is now,--neither more nor less.

[2489] _Records of Walmer_, p. 5.

[2490] _Coast Erosion_, p. 3.

[2491] _Report of ... the Brit. Association_, 1888 (1889), p. 910. The following table, compiled by Major A. C. Hepper, R.E. (_ib._, 1885, p. 440), illustrates the movements of the shingle during the period between 1741 and 1884:--

_Increase_ _Decrease_
_Place_ _From_ _To_ _Feet_ _Feet_
Walmer Castle 1741 1841 308 --
” ” 1841 1859 34 --
” ” 1859 1872 33 --
” ” 1872 1884 10 --
Deal Castle 1741 1859 85 --
” ” 1859 1872 -- 40
” ” 1872 1884 35 --
Sandown Castle 1741 1859 -- 145
” ” 1859 1872 -- 50
” ” 1872 1884 -- 5
No. 2 Battery 1859 1884 140 --

The encroachment of the sea north of Deal between 1848 and 1856 was due to the extraordinary prevalence of north-easterly winds.

[2492] _Britannia antiqua_, 1857, p. 282.

[2493] In regard to this statement, and also that of Roach Smith, recording the discovery of coins at Stonar (see p. 520, _supra_), Sir John Evans has written to me, ‘I have no personal knowledge of either of the finds of Roman coins that you mention. Roach Smith, however, and Beale Poste are competent authorities in such a case, and I see no reason why you should not accept their statements.’

[2494] _Archaeol. Cant._, xxv, 1902, p. 1.

[2495] _Ib._, pp. 4-5. This discovery stultifies Hasted’s remark (_Hist. of Kent_, iv, 1779, p. 173), that ‘towards the village of Walmer [as one comes from Deal] is a flat, many feet lower than the high-water mark, which the beach thrown up along the shore has fenced from the sea, and which probably when _Caesar_ landed on this coast might be all covered with water’. Cf. C. R. S. Elvin, _Records of Walmer_, p. 3.

[2496] See also _Archaeol. Cant._, xxvi, 1904, pp. 11-2.

[2497] Part i, 9th ed., 1900, p. 339.

[2498] _Britannia antiqua_, pp. 288-9.

[2499] _The Channel Pilot_, part i, 1900, p. 338.

[2500] _Geogr. Journal_, ix, 1897, p. 655.

[2501] G. B. Gattie, _Memorials of the Goodwin Sands_, 1890, pp. 3, 5-6.

[2502] _Principles of Geology_, 1875, i, 530-1.

[2503] _Anglo-Saxon Chronicle_, ed. B. Thorpe, ii, 1861, p. 203.--An. M.XCIX.

[2504] _Archaeol. Journal_, xxi, 1864, pp. 235-6.

[2505] _Treatise of the Roman Ports and Forts in Kent_, 1693, p. 24.

[2506] _Itin._, 1744, vii, 113.

[2507] _Villare Cantianum_, 1669, map facing p. 1.

[2508] See the fantastic map inserted between pages 330 and 331 of Guest’s _Origines Celticae_, vol. ii, in which ‘Lomea’ is placed N.W. of the Goodwins.

[2509] _Report of ... the Brit. Archaeol. Association ... _, Sept., 1844, p. 371.

[2510] _De rebus Albionicis_, 1590, pp. 24, 27-8.

[2511] According to Chambers’s _Ency._, v, 1901, p. 296, Lomea has been identified with ‘_Infera insula_ of the Romans’. The writer does not inform us by whom _infera insula_ was mentioned.

[2512] Part i, 1900, p. 337.

[2513] _Treatise of the Roman Ports and Forts of Kent_, pp. 20-1.

[2514] _Hist. Maps of England_, p. 2.

[2515] Richard Lilburne (_Topographie ... of ... Kent_, 1659, pp. 262-3), alluding to the well-known legend as to the origin of the Goodwin Sands, says, ‘the most probable relation of the rise of the same is thus. _Goodwin_ ... was ... owner of a great quantity of flat Lands in the County (neer the _Isle_ of _Thanet_) defended from the sea by a great wall, which lands afterwards (in the year 1099) was parcell of the possessions of the Abbot of St. _Augustine_ (but reteyned the name of _Goodwin_ ...), and that Abbot, being then also owner of the Rectory of _Tenterden_, and having begun the building of this steeple ... the thoughts, and actions, of him, and his agents were so set upon the finishing of that work, that they neglected the care of watching, and preserving the aforesaid wall, and (3. of November in that year) the sea broke over, and ... drowned the aforesaid lands (overwhelming the same) with a light sand ... and the place thereby obteyned the name of _Goodwin Sands_ ... and thus (accidentially) this _Tenterden_ steeple is said to be the cause of _Goodwin Sands_.’

[2516] _Hist. of Deal_, 1864, p. 106.

[2517] See p. 524, _supra_.

[2518] _Coast Erosion_, 1899, p. 12.

[2519] _Archaeol. Journal_, xlii, 1885, pp. 284-5. According to Mr. Clement Reid (_Archaeologia_, part ii, 1906, p. 285) ‘the relative level of sea and land in the south of England appears to have remained unchanged’ since ‘late Neolithic times’. See, however, Addenda, p. 740.

[2520] There is not much force in Professor Boyd Dawkins’s argument (_Early Man in Britain_, p. 483), that an island on the site of the Goodwin Sands would not have escaped the notice of Ptolemy. Ptolemy does not mention Sheppey (or else Thanet) and other islands.

[2521] See pp. 657-9, _infra_, and cf. R. Blanchard, _La Flamande_, pp. 128, 133.

[2522] _The Cinque Ports_, p. 8.

[2523] _Twenty-third Report East Kent Nat. Hist. Soc._, 1881, p. 57.

[2524] _Nouveau Dict. de Géogr. univ._, ii, 1884, p. 542.

[2525] According to M. Léon Lejeal, the author of an interesting article on ‘Le littoral’ in _Boulogne-sur-mer et la région boulonnaise_ (i, 365), ‘certains hydrographes affirment qu’au Grisnez, la falaise s’entame de 0,25 centimètres par an.’ I presume that this was the authority upon which M. de St.-Martin relied.

[2526] _Archaeol. Journal_, xxxiii, 1876, p. 60.

[2527] _Naut. Mag._, 1850, p. 216.

[2528] In Capt. McDakin’s _Coast Erosion,--Dover Cliffs_, 1899, pp. 7-9, a list is given of the notable falls which have been recorded. In 1853 there was a heavy fall near Holy Trinity Church, Dover; in 1872 at the East Cliff; in 1896 at the South Foreland; and (_Times_, Jan. 11, 1905, p. 7, col. 1, Jan. 13, p. 7, col. 2) in 1905 there were landslips at St. Margaret’s Bay, near Hope Point, and at Fan Bay.

[2529] Dowker (_Twenty-third Report East Kent Nat. Hist. Soc._, 1881, p. 63) attributed this loss of shingle to the Admiralty Pier at Dover. ‘The formation of the Dover Harbour,’ he says, ‘has favoured the accumulation of beach west of that point; the current, moreover, after passing the obstacle, is deflected inland, and thus, at St. Margaret’s Bay, a former collection of beach is being removed towards Deal.’ On the other hand, Sir John Coode, who is described in the _Dictionary of National Biography_ (Suppl., ii, 52) as ‘probably the most distinguished harbour engineer of the nineteenth century’, states (_Parl. Papers_, lviii, 1873, p. 455[3]) that ‘so far from the pier having acted as a check to the passage of the shingle, there has been a considerable loss to the westward of it within the last 20 years’. ‘I have no hesitation,’ he adds (_ib._, p. 456[4]), ‘in stating, in the most distinct and positive terms, that this decrease [of shingle on various parts of the coast south-west of St. Margaret’s Bay] has not been caused by “the extension of the Admiralty Pier at Dover”, inasmuch as the various facts that have been brought out in the course of my recent investigation lead distinctly and unmistakably to the opposite conclusion ... having regard to the facts previously stated, as to the diminution of shingle to the westward of Folkestone, near Sandgate and Hythe, &c. ... I have arrived at the conclusion that this [decrease of shingle between Dover and St. Margaret’s Bay] is due to the remarkable accumulation of shingle, and consequent projection towards the south-east of Dungeness’ (_ib._, p. 457[5]).

About the year 1721 the supply of shingle was temporarily cut off by the fall of part of the Castle Cliff. See Capt. John Perry, _Account of the Stopping of Daggenham Breach_, &c., 1721, p. 119.

[2530] _Report of ... the Brit. Association_, 1885 (1886), p. 439. See also pp. 406-7. According to J. B. Redman (_Proc. Inst. Civ. Engineers_, xi, 1851-2 [1853], p. 164) ‘it appears that at an early period there was no shingle at all at Dover ... which there is historical evidence to prove was the case; its gradual advance from the westward eventually blocked up the entrance’, &c. Where the ‘historical evidence’ is to be found Redman omits to say; and I cannot find it; but it is certain that the movement of shingle along the coast began long before the historic period (_Geogr. Journal_, xxviii, 1906, p. 489).

Capt. McDakin (_Coast Erosion,--Dover Cliffs_, p. 5) remarks that ‘the Roman Pharos on the Castle Cliffs and the foundations of a similar building in the Redoubt on the Western Heights, give us no indication that the edge of the cliff has receded since those earliest of Roman buildings occupied their present site’.

[2531] Clement Reid in _Vict. Hist. of ... Sussex_, i, 25, and _Geogr. Journal_, xxviii, 1906, pp. 488-9.

[2532] _Vict. Hist. of ... Sussex_, i, 469.

[2533] See _Geogr. Journal_, xxviii, 1906, p. 490.

[2534] _Ib._, p. 489. Cf. A. J. Jukes-Browne, _Handbook of Phys. Geol._, 1892, p. 171.

[2535] Angusti montes (_B. G._, iv, 23, § 3).

[2536] See p. 329, _supra_.

[2537] Experiments recently conducted by Captain McDakin (_Coast Erosion,--Dover Cliffs_, pp. 3-4, 12) showed that ‘the average erosion of four years was unexpectedly small, only amounting to half an inch in a year’. He admits, indeed, that the average rate, since erosion began, ‘has probably been much more rapid.’ His general conclusions are, ‘that the heaviest falls ... take place after long continued rain.... That the springs issuing from the base of the cliffs play an important part in undermining and bringing down the cliffs; and that the sea charged with a small amount of shingle [which it discharges like a gun] attacks the undercliff and removes it, but where the shingle accumulates in large quantities, it defends and supports the base of the cliffs,’ &c.

[2538] Ed. Wesseling, p. 473. See also _Corpus inscr. Lat._, vii, 1228.

[2539] _Hist. de Jules César_, ii, 157, note.

[2540] _Domesday Book of Kent_, ed. L. B. Larking, 1869, p. 93, and Extension, p. 2.

[2541] I need hardly say that Digges’s statement, which refers only to the inlet where the port of Dover had been, does not support Redman’s assertion (p. 529, n. 4, _supra_).

[2542] _Archaeologia_, xi, 1792, p. 212, note _a_. Archcliff Fort is about 400 yards west of the first groyne on the western side of the Lord Warden Hotel.

[2543] _Naut. Mag._, 1850, p. 269. See also John Leland, _Itin._, vii, 1744, fol. 128, p. 117.

[2544] About 1 mile 4,100 feet in a straight line from the present high-water mark of ordinary tides (_Six-inch Ordnance Survey_, Sheet 68).

[2545] Nearly 2 miles beyond Crabble (_ib._, Sheets 67-8).

[2546] _Archaeol. Cant._, xx, 1893, p. 129.

[2547] _Archaeologia_, v, 1779, p. 325; John Lyon, _Hist. of ... Dover_, i, 1813, p. 9; _Archaeol. Cant._, xx, 1893, p. 131.

[2548] _Ib._, xviii, 1889, p. 202.

[2549] See also T. Hyde Page, _Considerations upon the State of Dover Harbour_, &c., 1784, p. 6.

[2550] See _Phil. Trans._, xxix, 1716, p. 469; lxxvi, 1786, p. 220; W. Lambarde, _Perambulation of Kent_, ed. 1826, p. 154; _Mem. Geol. Survey,--The Geology of the Weald_, 1875, pp. 302, 315-6; _Proc. Geologists’ Association_, xiii, 1895, pp. 40-7; Capt. McDakin, _Coast Erosion,--Dover Cliffs_, pp. 7-9; _Pall Mall Gazette_, Jan. 18, 1906, p. 12, col. 2.

[2551] _Trans. Geol. Soc._, v, 1821, p. 17.

[2552] The quotation is from T. Lewin, _The Invasion of Britain by Julius Caesar_, 2nd ed., 1862, p. lvii.

[2553] Romney Marsh Proper extends eastward of the Rhee Wall, which runs from Appledore to New Romney.

[2554] _Mem. Geol. Survey,--The Geology of the Weald_, p. 251.

[2555] _Phil. Trans._, xxxv, 1727, pp. 551-2.

[2556] _The Invasion of Britain by Julius Caesar_, 1862, p. lii. We learn from the late F. Drew (_Mem. Geol. Survey,--The Geology of the Weald_, p. 206) that ‘whatever the soil may be near the surface, it is almost invariably the case that, at a depth of 10 or 20 feet, there is loose sand, often containing recent marine shells’, &c. See note 8, _infra_.

[2557] _Athenæum_, Aug. 5, 1865, pp. 184-5.

[2558] See _Geol. Mag._, 1869, p. 128. The writer, ‘W. T.,’ was evidently the late geologist, William Topley.

[2559] See _Proc. Geologists’ Association_, xv, 1898, pp. 212-3, 222. As far as I can discover, the only absolutely trustworthy boring which has been made (_Quart. Journ. Geol. Soc._, xliii, 1887, p. 204) shows that at Holmestone, near Lydd, which is outside the limits of Romney Marsh Proper, the recent strata, overlying Hastings beds, were as follows:--Shingle, 15 feet; Boulders, 4 feet; Brown Sand, 13 feet; Clay, 4 feet; Black and Grey Sand, 20 feet; Pebbles, 1 foot.

‘Mr. Elliott,’ says Drew (_Mem. Geol. Survey,--The Geology of the Country between Folkestone and Rye_, 1864, p. 16), ‘tells me that he bored 70 feet in the Marsh, of which the last 50 were in sand.’ ‘I contend, however,’ replies Dowker (_Proc. Geologists’ Association_, xv, 212), ‘that this does not prove anything, since the sand probably belonged to the Hastings Beds.’

[2560] _Mem. Geol. Survey,--The Geology of the Weald_, p. 304. Topley goes on to point out that F. H. Appach, in _C. J. Caesar’s Brit. Expeditions_, &c., p. 16, adopted a theory which had been originally put forward by James Elliott, but had been discovered by Elliott himself to be erroneous, attributing ‘the silting up of the area’ to ‘the presence of some supposed islands of Hastings Sand near Romney’.

[2561] _Mem. Geol. Survey,--The Geology of the Country between Folkestone and Rye_, pp. 19-20.

[2562] See p. 543, n. 1, _infra_.

[2563] See p. 62, _supra_.

[2564] See p. 543, _infra_.

[2565] See _Proc. Inst. Civil Engineers_, xl, 1875, pp. 69-70.

[2566] _Ib._, pp. 109, 111. Lord Avebury, who refers to this paper in _The Scenery of England_, 3rd ed., 1904, p. 152, reports Sir Joseph Prestwich’s views as to the movement of shingle incorrectly. Prestwich considered it ‘well established’ that the general movement of the shingle along our south coast was eastward, although in the west bay of Portland it travels in the opposite direction.

[2567] C. Roach Smith, _Report on Excavations ... at Lymne_, 1852, p. 41.

[2568] _Proc. Inst. Civil Engineers_, vi, 1847, p. 467.

[2569] T. Lewin, _The Invasion of Britain_, &c., 1862, pp. lxviii-lxix. See also p. lvi.

[2570] _Archaeologia_, xl, 1866, pp. 361-74.

[2571] Lewin observes (_The Invasion of Britain_, &c., 1862, pp. lvii-lviii) that as far eastward as West Hythe Oaks the shingle ‘fulls’ all curve westward, having been bent in that direction by the inrush of the tides; while from West Hythe Oaks to Sandgate they all curve towards the east. This, he says, proves that when they were formed, the mouth of the estuary near Hythe had already been closed. Appach, on the other hand (_C. J. Caesar’s Brit. Expeditions_, p. 21, § 9), does not believe that the shingle spit reached West Hythe Oaks. Referring to the change of curvature in the shingle fulls, he says that it was ‘evidently due to the cessation of the indraught’, which was ‘obviously caused by the erection of the ancient wall at West Hythe’. Hence, he concludes, ‘the fulls to the north of the point [where the change of curvature takes place] ... were not formed until after the wall at West Hythe was built; and as this is part of the north-eastern boundary of Romney Marsh, it follows that the fulls in question were formed after the formation of Romney Marsh.’ Lewin also mentions ‘the ancient wall at West Hythe’; but his final theory is that the erection of this dam became necessary because the shingle spit, after it had reached West Hythe Oaks, was burst by the waters, fed by the streams mentioned above (p. 532), which accumulated in the space between West Hythe Oaks and Hythe (see p. 547, _infra_). Appach holds that Romney Marsh was not formed until after the Romans had abandoned Britain; and he is therefore constrained to argue that Hythe Haven did not exist during the Roman occupation, and that the Portus Lemanis was at Lympne. Both of these theories will be refuted in this article (pp. 543-8, _infra_).