The Greeks who, after the Indian conquests of Alexander, brought back the earliest accounts of the East, repeated them without material correction, and reported the island to be nearly twenty times its actual extent. Onesicritus, a pilot of the expedition, a.s.signed to it a magnitude of 5000 stadia, equal to 500 geographical miles.[1]

Eratosthenes attempted to fix its position, but went so widely astray that his first (that is his most southern) parallel pa.s.sed through it and the "Cinnamon Land," the _Regio Cinnamomifera_, on the east coast of Africa.[2] He placed Ceylon at the distance of seven days' sail from the south of India, and he too a.s.signed to its western coast an extent of 5000 stadia.[3] Both those authorities are quoted by Strabo, who says that the size of Taprobane was not less than that of Britain.[4]

[Footnote 1: STRABO, lib. v. Artemidorus (100 B.C.), quoted by Stepha.n.u.s of Byzantium, gives to Ceylon a length of 7000 stadia and a breadth of 500.]

[Footnote 2: STRABO, lib. ii. c. i. s. 14.]

[Footnote 3: The text of Strabo showing this measure makes it in some places 8000 (Strabo, lib. v.); and Pliny, quoting Eratosthenes, makes it 7000.]

[Footnote 4: STRABO, lib. ii. c. v. s. 32. Aristotle appears to have had more correct information, and says Ceylon was not so large as Britain.--_De Mundo_ ch. iii.]

The round numbers employed by those authors, and by the Greek geographers generally, who borrow from them, serve to show that their knowledge was merely collected from rumours; and that in all probability they were indebted for their information to the stories of Arabian or Hindu sailors returning from the Eastern seas.

Pliny learned from the Singhalese Amba.s.sador who visited Rome in the reign of Claudius, that the breadth of Ceylon was 10,000 stadia from west to east; and Ptolemy fully developed the idea of his predecessors, that it lay opposite to the "Cinnamon Land," and a.s.signed to it a length from north to south of nearly _fifteen degrees_, with a breadth of _eleven_, an exaggeration of the truth nearly twenty-fold.[1]

Agathemerus copies Ptolemy; and the plain and sensible author of the "Periplus" (attributed to Arrian), still labouring with the delusion of the magnitude of Ceylon, makes it stretch almost to the opposite coast of Africa.[2]

[Footnote 1: PTOLEMY, lib. vii. c. 4.]

[Footnote 2: ARRIAN, _Periplus_, p. 35. Marcia.n.u.s Heracleota (whose Periplus has been reprinted by HUDSON, in the same collection from which I have made the reference to that of Arrian) gives to Ceylon a length of 9500 stadia with a breadth of 7500.--MAR. HER. p. 26.]

These extravagant ideas of the magnitude of Ceylon were not entirely removed till many centuries later. The Arabian geographers, Ma.s.soudi, Edrisi, and Aboulfeda, had no accurate data by which to correct the errors of their Greek predecessors. The maps of the fourteenth and fifteenth centuries repeated their distortions[1]; and Marco Polo, in the fourteenth century, who gives the island the usual exaggerated dimensions, yet informs us that it is now but one half the size it had been at a former period, the rest having been engulfed by the sea.[2]

[Footnote 1: For an account of Ceylon as it is figured in the _Mappe-mondes_ of the Middle Ages, see the _Essai_ of the VICOMTE DE SANTAREM, _Sur la Cosmographie et Cartographie_, tom. iii. p. 335, &c.]

[Footnote 2: MARCO POLO, p. 2, c. 148. A later authority than Marco Polo, PORCACCHI, in his _Isolario_, or "Description of the most celebrated Islands in the World," which was published at Venice in A.D.

1576, laments his inability even at that time to obtain any authentic information as to the boundaries and dimensions of Ceylon; and, relying on the representations of the Moors, who then carried on an active trade around its coasts, he describes it as lying under the equinoctial line, and possessing a circuit of 2100 miles. "Ella gira di circuito, secondo il calcole fatto da Mori, che modernamente l'hanno nauigato d'ogn'intorno due mila et cento miglia et corre maestro e sirocco; et per il mezo d'essa pa.s.sa la linea equinottiale et e el principio del primo clima al terzo paralello."--_L'Isole piu Famose del Monde, descritte da_ THOMASO PORCACCHI, lib. iii. p. 30.]

Such was the uncertainty thrown over the geography of the island by erroneous and conflicting accounts, that grave doubts came to be entertained of its ident.i.ty, and from the fourteenth century, when the attention of Europe was re-directed to the nascent science of geography, down to the close of the seventeenth, it remained a question whether Ceylon or Sumatra was the Taprobane of the Greeks.[1]

[Footnote 1: GIBBON states, that "Salmasius and most of the ancients confound the islands of Ceylon and Sumatra."--_Decl. and Fall_ ch. xl.

This is a mistake. Saumaise was one of those who maintained a correct opinion; and, as regards the "ancients," they had very little knowledge of _Further India_ to which Sumatra belongs; but so long as Greek and Roman literature maintained their influence, no question was raised as to the ident.i.ty of Ceylon and Taprobane. Even in the sixth century Cosmas Indicopleustes declares unhesitatingly that the Sielediva of the Indians was the Taprobane of the Greeks.

It was only on emerging from the general ignorance of the Middle Ages that the doubt was first promulgated. In the Catalan Map of A.D. 1375, ent.i.tled _Image du Monde_, Ceylon is omitted, and Taprobane is represented by Sumatra (MALTE BRUN, _Hist. de Geogr._ vol. i, p. 318); in that of _Fra Mauro_, the Venetian monk, A.D. 1458, Seylan is given, but _Taprobane_ is added over _Sumatra_. A similar error appears in the _Mappe-monde,_ by RUYCH, in the Ptolemy of A.D. 1508, and in the writings of the geographers of the sixteenth century, GEMMA FRISIUS, SEBASTIAN MUNSTER, RAMUSIO, JUL. SCALIGER, ORTELIUS, and MERCATOR. The same view was adopted by the Venetian NICOLA DI CONTI, in the first half of the fifteenth century, by the Florentine ANDREA CORSALI, MAXIMILIa.n.u.s TRANSYLVa.n.u.s, VARTHEMA, and PIGAFETTA. The chief cause of this perplexity was, no doubt, the difficulty of reconciling the actual position and size of Ceylon with the dimensions and position a.s.signed to it by Strabo and Ptolemy, the latter of whom, by an error which is elsewhere explained, extended the boundary of the island far to the east of its actual site. But there was a large body of men who rejected the claim of Sumatra, and DE BARROS, SALMASIUS, BOCHART CLUVERIUS, CELLARIUS, ISAAC VOSSIUS and others, maintained the t.i.tle of Ceylon. A _Mappe-monde_ of A.D. 1417, preserved in the Pitti Palace at Florence compromises the dispute by designating Sumatra _Taprobane Major_. The controversy came to an end at the beginning of the eighteenth century, when the overpowering authority of DELISLE resolved the doubt, and confirmed the modern Ceylon as the Taprobane of antiquity. WILFORD, in the _Asiatic Researches_ (vol. x. p. 140), still clung to the opposite opinion, and KANT undertook to prove that Taprobane was Madagascar.]

_Lat.i.tude and Longitude_.--There has. .h.i.therto been considerable uncertainty as to the position a.s.signed to Ceylon in the various maps and geographical notices of the island: these have been corrected by more recent observations, and its true place has been ascertained to be between 5 55' and 9 51' north lat.i.tude, and 79 41' 40" and 81 54'

50" east longitude. Its extreme length from north to south, from Point Palmyra to Dondera Head, is 271-1/2 miles; its greatest width 137-1/2 miles, from Colombo on the west coast to Sangemankande on the east; and its area, including its dependent islands, 25,742 miles, or about one-sixth smaller than Ireland.[1]

[Footnote 1: Down to a very recent period no British colony was more imperfectly surveyed and mapped than Ceylon; but since the recent publication by Arrowsmith of the great map by General Fraser, the reproach has been withdrawn, and no dependency of the Crown is more richly provided in this particular. In the map of Schneider, the Government engineer in 1813, two-thirds of the Kandyan Kingdom are a blank; and in that of the Society for the Diffusion of Knowledge, re-published so late as 1852, the rich districts of Neuera-kalawa and the Wanny, in which there are innumerable villages (and scarcely a hill), are marked as "_unknown mountainous region_." General Fraser, after the devotion of a lifetime to the labour, has produced a survey which, in extent and minuteness of detail, stands unrivalled. In this great work he had the co-operation of Major Skinner and of Captain Gallwey, and to these two gentlemen the public are indebted for the greater portion of the field-work and the trigonometrical operations. To judge of the difficulties which beset such an undertaking, it must be borne in mind that till very recently travelling in the interior of Ceylon was all but impracticable, in a country unopened even by bridle roads, across unbridged rivers, over mountains never trod by the foot of a European, and amidst precipices inaccessible to all but the most courageous and prudent. Add to this that the country is densely covered with forest and jungle, with trees a hundred feet high, from which here and there the branches had to be cleared to obtain a sight of the signal stations. The triangulation was carried on amidst privations, discomfort, and pestilence, which frequently prostrated the whole party, and forced their attendants to desert them rather than encounter such hardships and peril. The materials collected by the colleagues of General Fraser under these discouragements have been worked up by him with consummate skill and perseverance. The base line, five and a quarter miles in length, was measured in 1845 in the cinnamon plantation at Kaderani, to the north of Colombo, and its extremities are still marked by two towers, which it was necessary to raise to the height of one hundred feet, to enable them to be discerned above the surrounding forests. These it is to be hoped will be carefully kept from decay, as they may again be called into requisition.

As regards the sea line of Ceylon, an admirable chart of the West coast, from Adam's Bridge to Dondera Head, has been published by the East India Company from a survey in 1845. But information is sadly wanted as to the East and North, of which no accurate charts exist, except of a few unconnected points, such as the harbour of Trincomalie.]

_General Form_.--In its general outline the island resembles a pear--and suggests to its admiring inhabitants the figure of those pearls which from their elongated form are suspended from the tapering end. When originally upheaved above the ocean its shape was in all probability nearly circular, with a prolongation in the direction of north-east. The mountain zone in the south, covering an area of about 4212 miles[1], may then have formed the largest proportion of its entire area--and the belt of low lands, known as the Maritime Provinces, consists to a great extent of soil from the disintegration of the gneiss, detritus from the hills, alluvium carried down the rivers, and marine deposits gradually collected on the sh.o.r.e. But in addition to these, the land has for ages been slowly rising from the sea, and terraces abounding in marine sh.e.l.ls imbedded in agglutinated sand occur in situations far above high-water mark. Immediately inland from Point de Galle, the surface soil rests on a stratum of decomposing coral; and sea sh.e.l.ls are found at a considerable distance from the sh.o.r.e. Further north at Madampe, between Chilaw and Negombo, the sh.e.l.ls of pearl oysters and other bivalves are turned up by the plough more than ten miles from the sea.

[Ill.u.s.tration]

[Footnote 1: This includes not only the lofty mountains suitable for the cultivation of coffee, but the lower ranges and spurs which connect them with the maritime plains.]

These recent formations present themselves in a still more striking form in the north of the island, the greater portion of which may be regarded as the conjoint production of the coral polypi, and the currents, which for the greater portion of the year set impetuously towards the south.

Coming laden with alluvial matter collected along the coast of Coromandel, and meeting with obstacles south of Point Calimere, they have deposited their burthens on the coral reefs round Point Pedro; and these gradually raised above the sea-level, and covered deeply by sand drifts, have formed the peninsula of Jaffna and the plains that trend westward till they unite with the narrow causeway of Adam's Bridge--itself raised by the same agencies, and annually added to by the influences of the tides and monsoons.[1]

[Footnote 1: The barrier known as Adam's Bridge, which obstructs the navigation of the channel between Ceylon and Ramnad, consists of several parallel ledges of conglomerate and sandstone, hard at the surface, and growing coa.r.s.e and soft as it descends till it rests on a bank of sand, apparently acc.u.mulated by the influence of the currents at the change of the monsoons. See an _Essay_ by Captain STEWART _on the Paumbem Pa.s.sage_. Colombo, 1837. See Vol. II. p. 554.]

On the north-west side of the island, where the currents are checked by the obstruction of Adam's Bridge, and still water prevails in the Gulf of Manaar, these deposits have been profusely heaped, and the low sandy plains have been proportionally extended; whilst on the south and east, where the current sweeps unimpeded along the coast, the line of the sh.o.r.e is bold and occasionally rocky.

This explanation of the accretion and rising of the land is somewhat opposed to the popular belief that Ceylon was torn from the main land of India[1] by a convulsion, during which the Gulf of Manaar and the narrow channel at Paumbam were formed by the submersion of the adjacent land.

The two theories might be reconciled by supposing the sinking to have occurred at an early period, and to have been followed by the uprising still in progress. But on a closer examination of the structure and direction of the mountain system of Ceylon, it exhibits no traces of submersion. It seems erroneous to regard it as a prolongation of the Indian chains; it lies far to the east of the line formed by the Ghauts on either side of the peninsula, and any affinity which it exhibits is rather with the equatorial direction of the intersecting ranges of the Nilgherries and the Vindhya. In their geological elements there is, doubtless, a similarity between the southern extremity of India and the elevated portions of Ceylon; but there are also many important particulars in which their specific differences are irreconcilable with the conjecture of previous continuity. In the north of Ceylon there is a marked preponderance of aqueous strata, which are comparatively rare in the vicinity of Cape Comorin; and whilst the rocks of the former are entirely dest.i.tute of organic remains[2]; fossils, both terrestrial and pelagic, have been found in the Eastern Ghauts, and sandstone, in some instances, overlays the primary rocks which compose them. The rich and black soil to the south of the Nilgherries presents a strong contrast to the red and sandy earth of the opposite coast; and both in the flora and fauna of the island there are exceptional peculiarities which suggest a distinction between it and the Indian continent.

[Footnote 1: La.s.sEN, _Indische Alterthumskunde_, vol. i. p. 193.]

[Footnote 2: At Cutchavelly, north of Trincomalie, there exists a bed of calcareous clay, in which sh.e.l.ls and crustaceans are found in a semi-fossilised state; but they are all of recent species, princ.i.p.ally _Macrophthalmus_ and _Scylla_. The breccia at Jaffna contains recent sh.e.l.ls, as does also the arenaceous strata on the western coast of Manaar and in the neighbourhood of Galle. The existence of the fossilised crustaceans in the north of Ceylon was known to the early Arabian navigators. Abou-zeyd describes them as, "Un animal de mer qui resemble a l'ecrevisse; quand cet animal sort de la mer, _il se convert.i.t en pierre_." See REINAUD, _Voyages faits par les Arabes_, vol.

i. p. 21. The Arabs then; and the Chinese at the present day, use these petrifactions when powdered as a specific for diseases of the eye.]

_Mountain System_.--At whatever period the mountains of Ceylon may have been raised, the centre of maximum energy must have been in the vicinity of Adam's Peak, the group immediately surrounding which has thus acquired an elevation of from six to eight thousand feet above the sea.[1] The uplifting force seems to have been exerted from south-west to north-east; and although there is much confusion in many of the intersecting ridges, the lower ranges, especially those to the south and west of Adam's Peak, from Saffragam to Ambogammoa, manifest a remarkable tendency to run in parallel ridges in a direction from south-east to north-west.

[Footnote 1: The following are the heights of a few of the most remarkable places:--

Pedrotallagalla 8280 English feet.

Kirrigalpotta 7810 English feet.

Totapella 7720 English feet.

Adam's Peak 7420 English feet.

Nammoone-Koolle 6740 English feet.

Plain of Neuera-ellia 6210 English feet.]

Towards the north, on the contrary, the offsets of the mountain system, with the exception of those which stretch towards Trincomalie, radiate to short distances in various directions, and speedily sink down to the level of the plain. Detached hills of great alt.i.tude are rare, the most celebrated being that of Mihintala, which overlooks the sacred city of Anaraj.a.poora: and Sigiri is the only example in Ceylon of those solitary acclivities, which form so remarkable a feature in the table-land of the Dekkan, starting abruptly from the plain with scarped and perpendicular sides, and converted by the Indians into strongholds, accessible only by precipitous pathways, or steps hewn in the solid rock.

The crest of the Ceylon mountains is of stratified crystalline rock, especially gneiss, with extensive veins of quartz, and through this the granite has been everywhere intruded, distorting the riven strata, and tilting them at all angles to the horizon. Hence at the abrupt terminations of some of the chains in the district of Saffragam, plutonic rocks are seen mingled with the dislocated gneiss. Basalt makes its appearance both at Galle and Trincomalie. In one place to the east of Pettigalle-Kanda, the rocks have been broken up in such confusion as to resemble the effect of volcanic action--huge ma.s.ses overhang each other like suddenly-cooled lava; and Dr. Gygax, a Swiss mineralogist, who was employed by the Government in 1847 to examine and report on the mineral resources of the district, stated, on his return, that having seen the volcanoes of the Azores, he found a "strange similarity at this spot to one of the semi-craters round the trachytic ridge of Seticidadas, in the island of St. Michael."[1]

[Footnote 1: Beyond the very slightest symptoms of disturbance, earthquakes are unknown in Ceylon: and although its geology exhibits little evidence of volcanic action (with the exception of the basalt, which occasionally presents an appearance approaching to that of lava), there are some other incidents that seem to suggest the vicinity of fire; more particularly the occurrence of springs of high temperature, one at Badulla, one at Kitool, near Bintenne, another near Yavi Ooto, in the Veddah country, and a fourth at Cannea, near Trincomalie. I have heard of another near the Patipal Aar south of Batticaloa. The water in each is so pure and free from salts that the natives make use of it for all domestic purposes. Dr. Davy adverts to another indication of volcanic agency in the sudden and profound depth of the n.o.ble harbour at Trincomalie, which even close by the beach is said to have been hitherto unfathomed.

The Spaniards believed Ceylon to be volcanic; and ARGENSOLA, in his _Conquista de las Malucas_, Madrid, 1609, says it produced liquid bitumen and sulphur:--"Fuentes de betun liquido y bolcanes de perpetuas llamas que arrojan entre las asperezas de la montana losas de acufre."--Lib. v. p. 184. It is needless to say that this is altogether imaginary.]

_Gneiss_.--The great geological feature of the island is, however, the profusion of gneiss, and the various new forms arising from its disintegration. In the mountains, with the exception of occasional beds of dolomite, no more recent formations overlie it; from the period of its first upheaval, the gneiss has undergone no second submersion, and the soil which covers it in these lofty alt.i.tudes is formed almost entirely by its decay.

In the lower ranges of the hills, gigantic portions of gneiss rise conspicuously, so detached from the original chain and so rounded by the action of the atmosphere, aided by their concentric lamellation, that but for their prodigious dimensions, they might be regarded as boulders.

Close under one of these cylindrical ma.s.ses, 600 feet in height, and upwards of three miles in length, the town of Kornegalle, one of the ancient capitals of the island, has been built; and the great temple of Dambool, the most remarkable Buddhist edifice in Ceylon, is constructed under the hollow edge of another, its gilded roof being formed by the inverted arch of the natural stone. The tendency of the gneiss to a.s.sume these concentric and almost circular forms has been taken advantage of for this purpose by the Singhalese priests, and some of their most venerated temples are to be found under the shadow of the overarching strata, to the imperishable nature of which the priests point as symbolical of the eternal duration of their faith.[1]

[Footnote 1: The concentric lamellar strata of the gneiss sometimes extend with a radius so prolonged that slabs may be cut from them and used in subst.i.tution for beams of timber, and as such they are frequently employed in the construction of Buddhist temples. At Piagalla, on the road between Galle and Colombo, within about four miles of Caltura, there is a gneiss hill of this description on which a temple has been so erected. In this particular rock the garnets usually found in gneiss are replaced by rubies, and nothing can exceed the beauty of the hand-specimens procurable from a quarry close to the high road on the landward side; in which, however, the gems are in every case reduced to splinters.]

_Laterite or "Cabook_."--A peculiarity, which is one of the first to strike a stranger who lands at Galle or Colombo, is the bright red colour of the streets and roads, contrasting vividly with the verdure of the trees, and the ubiquity of the fine red dust which penetrates every crevice and imparts its own tint to every neglected article. Natives resident in these localities are easily recognisable elsewhere, by the general hue of their dress. This is occasioned by the prevalence along the western coast of _laterite_, or, as the Singhalese call it, _cabook_, a product of disintegrated gneiss, which being subjected to detrition communicates its hue to the soil.[1]

[Footnote 1: According to the _Mahawanso_ "Tamba-panni," one of those names by which Ceylon was anciently called, originated in an incident connected with the invasion of Wijayo, B.C. 543, whose followers, "exhausted by sea-sickness and faint from weakness, sat down at the spot where they had landed out of the vessels, supporting themselves on the palms of their hands pressed to the ground, whence the name of Tamba-pannyo, '_copper-palmed_,' from the colour of the soil. From this circ.u.mstance that wilderness obtained the name of Tamba-panni; and from the same cause also this renowned land became celebrated under that name."--TURNOUR'S _Mahawanso_, ch. vi. p. 50. From Tamba-panni came the Greek name for Ceylon, _Taprobane_. Mr. de Alwis has corrected an error in this pa.s.sage of Mr. Turnour's translation; the word in the original, which he took for _Tamba-panniyo_, or "copper-palmed," being in reality _tamba-vanna_, or "copper-coloured." Colonel Forbes questions the accuracy of this derivation, and attributes the name to the _tamana_ trees; from the abundance of which he says many villages in Ceylon, as well as a district in southern India, have been similarly called.

(_Eleven Years in Ceylon_, vol. i. p. 10.) I have not succeeded in discovering what tree is designated by this name, nor does it occur in MOON'S _List of Ceylon Plants_. On the southern coast of India a river, which flows from the ghats to the sea, pa.s.sing Tinnevelly, is called Tambapanni. Tambapanni, as the designation of Ceylon, occurs in the inscription on the rock of Girnar in Guzerat, deciphered by Prinsep, containing an edict by Asoka relative to the medical administration of India for the relief both of man and beast, (_Asiat. Soc. Journ. Beng._ vol. vii. p. 158.)]

The transformation of gneiss into laterite in these localities has been attributed to the circ.u.mstance, that those sections of the rock which undergo transition exhibit grains of magnetic iron ore partially disseminated through them; and the phenomenon of the conversion has been explained not by recurrence to the ordinary conception of mere weathering, which is inadequate, but to the theory of catalytic action, regard being had to the peculiarity of magnetic iron when viewed in its chemical formula.[1] The oxide of iron thus produced communicates its colouring to the laterite, and in proportion as felspar and hornblende abound in the gneiss, the cabook a.s.sumes respectively a white or yellow hue. So ostensible is the series of mutations, that in ordinary excavations there is no difficulty in tracing a continuous connection without definite lines of demarcation between the soil and the laterite on the one hand, and the laterite and gneiss rock on the other.[2]

[Footnote 1: From a paper read to the Royal Physical Society of Edinburgh by the Rev. J.G. Macvicar, D.D.]

[Footnote 2: From a paper on the Geology of Ceylon, by Dr. Gardner, in the Appendix to Lee's translation of RIBEYRO'S _History of Ceylon_, p, 206. The earliest and one of the ablest essays on the geological system and mineralogy of Ceylon will be found in DAVY'S _Account of the Interior of Ceylon_, London, 1821. It has, however, been corrected and enlarged by recent investigators.]