MALDIVE ISLANDS, a remarkable archipelago in the Indian Ocean, the northern extremity of which is 7" west of Ceylon, and which extends in length from north to south, from 7° 7' N. lat. to 0° 42' S., a space of 540 British miles (or about as far as from Kirkwall in Orkney to Dover), and is limited in width by the meridians 72° 27' and 73° 50'.[328-1] The strange appearance which this group assumes in the old maps of the 16th and 17th centuries (see fig. 2, from Mappemonde, cited on p. 329) is entirely inaccurate in detail, but hardly so singular as the reality exhibited by modern surveys.
The archipelago is in some respects one of the most dis-tinctly typical examples of a great aggregation of coral islands; indeed the technical name adopted by modern science for the annular coral formation which they exhibit (viz., atoll) has been taken from the language of these islands.[328-2] For Mr Darwin's theory of such formations see vol. vi. p. 378. Objections to this have recently been raised by Mr John Murray, but these do not affect the description.[328-3]
The Maldive archipelago in plan may be compared to a chain suspended from a peg, each link of which chain is an irregularly elliptical chaplet of islets, the greater axes of these quasi-ellipses varying from about 90 miles down-wards. Taking separately any one of these chaplets (or atolls), we now know it to be the nearly level summit of a submarine table-mountain, rising abruptly from the un-fathomable ocean, and approaching the surface within a distance which varies in different atolls from 20 to 45 fathoms. The quasi-elliptical margin of the atoll is fringed, and the central expanse of its area is more or less sparsely studded, " with oval basins of coral-rock just lipping the surface of the sea, and each containing a lake of clear water " (Darwin). These small oval basins, or ring-shaped reefs and islets, are in fact essentially miniatures of the atoll itself.
The general impression made by the Maldive atoll is vividly drawn by the French adventurer Pyrard de la Val (1602-7):
" Each atollon is detached, and contains within it a great multi-tude of small islands. It is a marvel to see one of these atollons, compassed all round by a great bank of stone, insomuch that no art of man could so well enclose with walls an equal space of ground Looking from the middle of an atollon you see all round you that great bank of stone encircling the isles and defend-ing them against the violence of the sea. And it is a fearful thing even for the boldest to draw near this bank and see the waves come on and break furiously all round .... so that you see all round you as it were a whitened wall."
Though the barrier reef, or banc de pierre, of which Pyrard speaks, exists in most of the atolls, there is none in the most northerly of the great atolls (Tiladummati and Milladummadu, two divisions of one atoll). In this there are broad and safe navigable channels, from 1 to 2 miles wide, between all the islands forming the chaplet. A vessel can enter the atoll by any one of these channels, and steer within it in any direction, anchoring anywhere on a sandy bottom in 20 to 25 fathoms. In the more southerly atolls entrance channels are only found at occa-sional intervals, though in all they are pretty numerous. Thus in Suadiva, the most southerly of the large atolls (50 miles from north to south, 36 miles from east to west), which has a barrier reef on great part of its contour, there are forty-two channels by which a ship can enter the lagoon.
It is observed that in the double part of the chain of atolls the openings are most numerous on those sides which are in juxtaposition. Thus on the three atolls of
Maldive Islands. Fig. 1, Admy. Chart; Fig. 2, Map of cir. 1555.
Ari and North and South Nilandu there are on the inner or eastern side seventy-three deep-water channels through the barrier, on the outer or western side only twenty-five; whilst on the atolb of South Male, Felidu, and Mulaku, which lie facing the three former, there are on the inner or western side fifty-six deep-water openings, and on the outer or eastern side only thirty-seven. These differences are doubt-less due to differences in the action of the sea caused by the juxtaposition of the two rows of atolls, and analogous facts are observed elsewhere in atolls exposed to trade-winds.
Immediately outside of the great chaplets or atolls the figures of the soundings rise suddenly. Thus at Ihavan-diffulu (Heawandoo of charts), the most northerly atoll, close to the margin of the reef the line gave 50 and GO •fathoms, and at 300 yards distance there was no bottom with a 300-fathom line. And this sudden increase of depth applies to the deep channels between the atolls as much as to the ocean east and west of them.
We have spoken of the small reefs, which fringe the atolls and dot their area, as also ring-shaped. This is the type, but it is not universal. The charts show that where the channels or breaches through the marginal reef of the atoll are few or narrow there are no minor annular reefs (e.g., in Suadiva); where the channels are somewhat broader, the intercepted portions of reef are annular, but • not the reef in the central area; where the channels are broadest, almost every reef throughout the atoll is more or less perfectly annular. The depth of the lagoon within these rings is generally 5 to 7 fathoms, but sometimes, as in Ari atoll, it reaches 12 fathoms. The outer margin of the rings is bordered with living coral, within which is a flat surface of coral-rock. On this flat, sand partially indurated, and fragments of coral, &c, have accumulated, and been converted into islets clothed with vegetation. Such islets sometimes fill the whole ring of reef, and some-times are mere strips occupying a segment of it. Obviously the whole aggregate of actual dry land in such an archipelago is infinitesimal compared with the area of the atolls. The highest part of the islands is generally about 6 feet above water. Moresby found the surface-sand usually about 3 feet thick, the upper part partially mixed with vegetable matter so as to form a light soil; below this a white compact sand, and then a soft sandstone 2 feet thick,, below which it softened to sand again, and fresh-water appeared.
All the islands of any extent are well clothed with wood, including many fine large trees and the ordinary shrubs of the Ceylon coast-jungle; where the jungle has been cleared, grass grows luxuriantly. But the cocoa-palm is the char-acteristic tree; and, low as the islands are, being covered with these, they can be seen from a masthead at 15 miles. The appearance they present is that of a tuft or line of trees rising out of the water.
A good deal of vicissitude seems to go on in the forma-tion of new islets and decay of old ones, of which our survey-officers met with various instances.
All the inhabited islands, and some besides, afford fresh water. But the quality of water varies; and it is not uncommon to see two wells within a few feet of each other, one brackish and the other excellent. None of the wells are more than 6 feet deep.
The whole archipelago has from the earliest reports of it formed a little kingdom. Physically the number of atolls may be reckoned as nineteen, besides some solitary islands ; bat administratively these are grouped into thirteen, and the term atoll has been transferred to this division. We give in the following table the list of these (political) atolls, in a second column the spelling of the marine charts, and in a third the list of atolls as given by Pyrard de la Val, in the beginning of the 17th century.
== TABLE ==
[329-1]
The list from Pyrard shows that the division in the beginning of the 17th century was identical with what it now is. But we may gather that it is substantially of much greater antiquity, from the statement of Ibn Batuta (c. 1343), who says the islands were divided into aklim (klimata, Gk.), each under a governor. He mentions eleven of these:—Balibur, Kannalus, Mahal, Tiladib, Karaidu, Tim, Tiladummati, Hiladummati, Baridu, Kandakal, Muluk,—of which indeed the names of only seven, viz., (1) Tiladummati, (2) Heladummati, (3) Balibiir, (5) Mahal, (7) Baridu, (8) Muluk, (12) Suweid, can be identified with those of the existing divisions. But another, Karaidu, no doubt repre-sents Karadiva, a well-known solitary island north of Male atoll; Kandakal is an island of the Miladummadu atoll, called in the charts Condaicoll; Tim appears near the north of Tilladummati as Oteim; and the three—Kannalfls, Kandakal, and Tim—are presented prominently as the islands Camdalus, Camdicall, and Otimo in the Mappemonde made for Henry II. of France (c. 1555, see Jomard's Facsimiles, livr*. vi., copied in fig. 2 supra; and compare Portolano of 1570, copied in Mr Birch's translation of Albuquerque's Commentaries). Possibly, therefore, the Moorish traveller had substituted true names of islands which he remembered for the names of atolls which he had forgotten.
The Maldives are inhabited by a people of old civiliza-tion, professing Islam, and ruled by a sultan of ancient lineage. What the number of islands may be we cannot say. They are popularly estimated at 12,000, as appears by the ancient style of the sultan as "king of 12,000 islands and 13 atolls." (See also Marco Polo, 2d ed., 1875, ii. 417-19.) Those marked with names in the British survey amount to 602, and the inhabited islands to 178. The men are of a darkish copper colour, short stature (5 feet 2 inches), and poor physique, but oval contour of face, pleasing expression, and large bright eyes, suggesting resemblance to both the Singhalese and Malabar people. The women are fairer than the men, with regular features and clean healthy aspect. A few of the people bear signs of African mixture, easily accounted for; and probably the blood of the small communities has been tinged by the occasional settlement of other foreigners. The people are decidedly unwarlike; and there is hardly any crime of violence among them. They are said to be lax in morals and conversation; but otherwise their character and disposi-tion have favourably impressed visitors. Though suspicious of strangers, they are hospitable; and among themselves they are kindly, and affectionate to their kindred and in attendance on the sick. They are very cleanly in person and domestic habits. The population has been guessed in some books at 200,000; almost certainly one-tenth of that number would be an ample estimate. Moresby states the population of 98 islands, and the aggregate is 11,310. In the same proportion 178 islands would give 20,543; but the aggregate quoted includes the Kiug's Island, which is much above the average in population.
The language is undoubtedly a dialect of Singhalese approaching the old Elu, but indicating a separation of ancient date, and it is more or less Mohammedanized. Nothing at present can be said of grammar. But Mr Albert Gray has drawn out in parallel columns the Maldivian words given by Pyrard with the modern Singhalese equi-valents (see Jour. Roy. As. Soc, quoted at end). A cursory analysis of the list (which contains 275 words) gives the following result:—
== TABLE ==
Combining 1, 2, and 7, we have 61 per cent, of words of Singhalese or Sanskrit origin. And an analysis by Mr Bell of one of the sultan's letters to the Ceylon Govern-ment gives 65 per cent, of such words.
The origin of the name Mal-dîva or Mâldîva is obscure. Dîva is a familiar word in the Indian prakrits (Sansk., dvîpa ; Pali, dîpo) for an island. By a form of this word the people formerly designated themselves and their country, and this survives in letters of last century from the sultan to the Ceylon Government, in which he designates his kingdom as Divehi Râjjé, and his subjects as Divehi mîhun, " island people." There is a very old example of this use in Ammianus Marcellinus, who, in reference to the alleged excitement in the East at the accession of Julian, says that missions were prepared " ab usque Divis et Serendivis," from the people of the Divas and of Serendîva or Ceylon. And this is the name Dîva or Dîba-jât (Pers. plural form) by which these islands are described by the early Arabian geographers. The first literary use of the whole name is Ibn Batata's Dhîbat-al-Mahal (14th century), an Arabized form, sometimes used (Mahal-dîb) by the people now, though the proper form seems to be Malé. Malé-diva may possibly, as Bishop Caldwell (comp. Grammar, 2d ed., Introd. p. 28) and others have suggested, have meant the "islands of Male" or Malabar. On the other hand Mâlâ (Sansk.), "a chaplet" or "row," is not an impossible etymology considering how naturally the word " chaplet" occurs in the endeavour to describe an atoll. But these are conjectures. Under the sultan (who styles himself on coinage " Lord of Land and Sea ") there used to be six recognized viziera or councillors (but this system is now obsolete), besides a chief of law and religion called fandiari. Over each of the thirteen atolls is a king's agent, called atolu-veri, who collects the revenue. This official is often one of the royal family, or a vizier's son, and often resides at Mâlé, employing a deputy. On each island is a headman called rarhu-veri. There is also on each island containing forty inhabitants a kâtibu (Ar., kâtib, "scribe"), who acts as judge and minister, celebrat-ing marriages, &c. Pyrard calls him the curé.
Some of the oldest accounts of these islands represent them as always governed by a woman,—a notion which probably arose among the Mohammedan visitors from finding that female heirs were not precluded from succes-sion. Just the same notion was held about Achin in the 17th century, because there chanced to reign there several female sovereigns in succession (see vol. i. p. 97). We do find females nominally reigning on the Maldives on two-of the rare occasions when we have glimpses of their state, viz., in the time of Ibn Batuta, and again in the last century.
Islam is universally professed by the people, nor is there-tradition of any other religion, though there are a variety of Pagan superstitions and some doubtful traces of Buddhism. Thus the Bo-tree (or pippal), so sacred amoug the Buddhists of Ceylon, is still cherished near mosques. Pyrard de la Val was informed that the conversion to Islam took place two centuries at most before his time, i.e., about 1400. But, unless there was a decay and revival, we know this to be wrong, as the islanders were Mohammedan in the time of Ibn Batuta (1343). And this traveller tells that the father of one of his wives in the islands had for his grandfather (though the word used may mean " ancestor" only) the Sultan Daud, who was grandson of Ahmed Shanil-raza, the first king who adopted Islam. Accepting the meaning of " grandfather," this would carry the con-version back to about 1200, a probable epoch, for about that time there was a considerable outburst of missionary zeal in Islam, which led to the conversion of the coast states of Sumatra, &c. Ibn Batuta records an inscription on the Jami Mosque of the King's Island which ran:— " Sultan Ahmed Shanu-razah embraced Islam at the hands of Abu'l Barakat the Berber from the West"; but no date is given.
We have mentioned the occurrence of the name Divi in Ammianus. At an earlier date Ptolemy notices the numerous islands lying in front of Taprobane, alleged to number 1378. It is possible also that the Mamiolae of the same geographer may constitute a duplicate indication of the Maldives. For in the gossip of Palladius about India (see C. Miiller's Pseudo-Callisthenes, p. 102) this name Mamiolae, is applied to a gi-oup of islands, 1000 in number, that lay near Taprobane, and respecting which the old fable of the magnetic rock was current, which Ptolemy also connects with the Mamiolae.
Cosmas (c. 545) shows distinct knowledge of the Maldives (with-out naming them) as numerous close-set small islands round Sielediba or Taprobane, in all of which were found cocoa-nuts and fresh water. Passing next to the Arabian notices translated by Renaudot and Reinaud, which date from 851, and to the work of Mas'udi in the next century, we find tolerably correct accounts of the Dîbas, said to be 1900 in number. Al Biriini's account (c. 1030) of the islands is marked by his usual perspicacity and accuracy. The Dîvas are islets which form themselves in tlie sea, appearing like a ridge of sand, extending and uniting till they present a solid aspect. But also with time some decompose and melt away in the sea, whilst the inhabitants transfer their cocoa-trees and possessions to an island which is waxing instead of waning,—circumstances corroborated by modern observation.
All the old authors speak of coir (the fibre of the cocoa-nut husk) as one of the staple products of these islands, and the importance of this article for marine equipment led the Portuguese about 1518 to establish a factory on the Maldives. Joao Gomez, the head of the settlement, was at first well received, but his arrogant and violent conduct gave great offence to the Mohammedan traders from Cambay, who brought an armed flotilla against the Portuguese and put them all to death. The Portuguese several times renewed the attempt to establish themselves on the islands, and maintained a garrison for some time, but these endeavours had no permanent result. The islanders were also frequently subject to raids at the hands of the Mopla pirates of Malabar, and sometimes also, it would seem, to maltreatment from the crews of European vessels. The MS. diary of Mr (afterwards Sir AVilliam) Hedges, who passed through the Maldives in 1685, says : '' We putt out a piece of a Eed Ancient to appear like a Moor's Vessell: not judging it safe to he known to be English ; Our Nation having lately gott an ill Name by abusing ye Inhabitants." Such circumstances probably led the islanders to place themselves in relation with the rulers of Ceylon; and in 1645 occurs the first record of the embassy from the sultan of the Maldives to the Dutch governor at Colombo, which has continued to the present day, under Dutch and English, to arrive annually, bringing some poor offering, as a vague token of homage and claim of protection. The last political trouble of which we have notice occurred in the middle of last century. In 1753 the chief minister conspired to hand over the islands to the Ali raja of Cannanore. A Mopla force occupied Male, and carried off the sultan. The traitor himself was rewarded by being thrown into the sea. The oppression of these foreigners made the islanders rise and expel them. The sultan never returned, and a minister who had ruled on his behalf assumed the kingdom in 1760. In 1754 Dupleix occupied Male with a small French detachment, which remained several years. In 18H the sultan wrote to the governor-general (Lord Minto) to complain of the violent conduct of the officers of a ship under British colours which had been wrecked on the islands. Lord Minto sent back a courteous answer with presents. There have been no other events during the British rule in Ceylon, and the last sultan, Mohammed Moidin, reigned without dispute from 1835 to 1882.
We have only three substantial accounts of the Maldives from actual residence:—(1) that of the Moor Ibn Batuta, who lived upon them more than a year (1343-44), and filled the office of cadi; it contains much curious detail; (2) the narrative of Francois Pyrard de la Val, a French adventurer on board a ship of St Malo, which was wrecked on a reef of the Malosmadulu atoll in 1602, and who was detained five years on the islands,—a book of the greatest interest and accuracy, and by far the best account of these islands in existence ; (3) a memoir by two officers of the Indian navy, Lieut. Young and Mr Christopher, who had been employed in the survey of the islands under Captain R. Moresby in 1834-35, and who volunteered to remain behind at Male, in order to acquire a knowledge of the language, customs, &c, of the inhabitants,— a laudable effort, but the result of it was marred somewhat by the illness which prostrated both officers.
The sultan's residence and the capital of the kingdom is the island of Male, which lies near the middle of the archipelago on the east side. It is about 1 mile long by f mile wide, and contains a population approaching 2000. It has been at one time encompassed with walls and bastions, but these continue in repair only on the north and west. On the north too is an old fort, apparently of Portuguese construction, with a few old guns. On the north and west sides also advantage has been taken of the encircling reef as the base of a wall which has been built up so as to form the lagoon into a harbour for small craft, having a depth of 6 to 12 feet, and a width of 150 yards. The town is laid out in long regular streets at right angles, shaded with trees; the houses are in " compounds," with high fences excluding the street, and are surrounded with fruit trees and flowers. The sultan's palace, a large upper-roomed house, occupies with its appurtenances an area of \ square mile, enclosed by a shallow ditch now choked with vegetation. The houses generally are large cottages of about 28 feet by 12, formed of substantial wooden frame, with peaked roofs thatched with cocoa-leaves; the walls are matted with cocoa-leaves, but sometimes planked. There are several mosques, and at least one minaret, about 40 feet high, for the call to prayer. Stone-built houses, common in Pyrard's time, are so no longer; there is now but one. There are marked distinctions of rank among the people. At least six classes (we hardly know whether to call them castes) are recognized, of whom the two highest form a pure aristocracy. The sixth class, called Kallo ("black" ?), consists of the common people generally, of whom the toddy-drawers are regarded as the lowest.
The employments of the common people are fishing, gathering cocoa-nuts and cowries, weaving, and toddy-draw-ing. Women beat the cocoa-fibre and twist it into yarn, make mats, prepare breadfruit by slicing and drying in the sun, spin and dye cotton thread, make sweetmeats of cocoa-nut and palm-sugar. Women are not secluded or veiled as in typical Moslem countries.
Rice, the staple of food, is imported. Other chief food is fish (chiefly dried bonito), breadfruit prepared in various ways, cocoa-nut, and a few fruits and vegetables. There are a few sheep and cattle on Malé island, which are oc-casionally slaughtered.
From the earliest notices the production of coir, the collection of cowries, and the weaving of excellent textures on these islands have been noted. This last, and that of fine mats are the only manufactures in which skill is shown. The mats seem to be now produced only in Suadiva atoll; the cloth chiefly, but not solely, in Malosmadulu atoll.
The chief exports of the islands, besides coir and cowries (a decreasing trade), are cocoa-nuts, copra (i.e., cocoa-nut husk), tortoise-shell, and dried bonito-fish. An enormous amount of this last was formerly carried to Ceylon and Sumatra, the latter being supplied by traders who came from Chittagong. It has been known over the East from time immemorial as koboli-mâs, a corruption apparently of the Maldivian kalû-bili-mâs, " black bonito fish," sometimes further corrupted to gomulmutch.
Native vessels of 80 to 200 tons burthen make annual trips to Calcutta towards the end of the south-west monsoon, returning with the north-east monsoon in December. After leaving the Maldives they sight no land till Jagannath. They carry thither the articles named above, and bring back rice, cotton stuffs, and sundries. These long voyages are not confined to the craft of the capital. Moresby, in 1834-35, found that a small island in the North Nilandu atoll sent annually to Bengal five or six boats of 80 to 100 tons each. On the same island there was a kind of navigation school, and the natives made and repaired some kind of nautical instruments. The old cash of the Maldives was the curious lârîn or " fish-hook money " made of a bent rod of silver. This has been long replaced by coins of base metal bearing the same name. The Anglo-Indian rupee is current for larger payments, and cowries are still used to some extent.
Two alphabets are known on the islands (besides the Arabic, which appears on tombstones and in other inscrip-tions). The first is an ancient alphabet, known as Divehi Hakura, " island letters." This in 1835 still survived in the southern atolls, and orders for these were written in it. It is written, like all the Indian alphabets, from left to right, and is evidently (by comparison with plate xvii. in Dr Burnell's Elements of South. Indian Palaeography) a form (with additional letters) of the old Tamil character (700 to 1300 A. D.) called in Malabar Vatteluttu, or "round hand." [331-1] The modern Maldive writing, called Gabali Tana, is usually [331-2] written from right to left, like Arabic. It is said to have been introduced in the 16th century, and has gone through several variations. Some of the letters are modified from the Arabic character, and nine of them are the Arabic numeral digits. On the other hand numerals are represented by letters of the alphabet. The former system of reckoning was duodecimal, but this is dying out.
Nothing is accurately known of the flora of the islands, and Kew possesses no illustration of it. Among larger trees are mentioned the banyan, pippal, breadfruit, tamarind, and a large tree called kandu, affording a very light wood used for rafts, floats, &c. ; also species of paudanus. The castor-oil tree is abundant, though not used. The cocoa-nut of the islands, though of fine quality, is very small, not much larger than an orange. The tree itself furnishes the only indigenous wood used for boat-building. The dumbari (Calophyllum inophyllum) and kuradi (Pemphis acidula) are used in minor wood-work. A tuber, grated and steeped in water to remove its acridity, is made into flour,—perhaps a Colocasia, which Ibn Batuta mentions (al-kalakas) as used to make a kind of vermicelli. They have also sweet potatoes, pine-apples, pomegranates (bearing fruit throughout the year), plantains, and most of the other tropical or subtropical Indian fruits, chillies, a few areca trees, &c. The double cocoa-nut of the Seychelles Islands (fruit of Lodoicea Sechellarum) used to be cast up on the islands, and was believed to be a submarine proauction,—hence called the sea cocoa-nut. It was valued for imaginary qualities, and exported to India. The Portuguese long believed it to be a product of these islands, and called it the Maldive cocoa-nut.
Animals are few. Those named are rats, numerous and destructive, which climb the cocoa-trees and devour the kernels; the large bat called in India "flying-fox," also said to destroy many small cocoa-nuts; tortoises; a small snake said to be harmless, &c. Domestic animals are rare; a few goats and cattle are reared on Male.
The climate is not oppressive or disagreeable, but is very unhealthy for strangers, whether Asiatic or European. Ibn Batuta says every visitor was attacked by violent fever; Pyrard says the same; and this was substantially the experience of the survey officers and crews in 1834-35. The native crews also suffered much from the disease called beri-beri (which has dropsical symptoms, and is often fatal) and from violent bowel-complaints.
A complete report on the Maldives has recently been prepared by Mr H. C. P. Bell of the Ceylon civil service, who has visited the islands, and this is now being printed at Colombo. Mr Bell kindly enabled the present writer to see a copy before this article went to press, and many valuable facts have been added from it. Other materials used have been—Darwin, The Structure and Distribution of Coral Reefs, 1842 ; Voyage de Francois Pyrard de la Val, Paris, 1679 (previous editions 1611, 1615-16, 1619); Voyages d'Ibn Batoutah, trans, of Defréméry and Sanguinetti, tom, iv., Paris, 1858 ; Hamilton, Desc. of Hindostán, ii. 299 ; Moresby, Naut. Directions for the Maldiva Islands, &c., 1840; Young and Christopher, in Trans. Bomb. Geog. Soc., vol. i. pp. 54-86 ; also see ibid. p. 102 and p. 313; Trans. Roy. Geog. Soc., vol. ii. pp. 72-93; also vol. v. p. 398; Jour. As. Soc. Bengal, vol. v. p. 794 ; Jour. Roy. As. Soc., vol. vi. pp. 42-76 ; ibid., new series, vol. x. pp. 173-209 (paper by Mr Albert Gray), &c. (H. Y.)
Footnotes
328-1 The solitary island of Minucoy (Minakai), lying 70 miles north of the Maldives (med. lat. 8° 16' 30" N.; population 3000), pertains to these islands by the race and language of its people, but, as it has long belonged to the raja of Cannanore, it is usually classed with the Laccadives.
328-2 Maid, atolu. The word atollon is already defined as a generic ex-pression in Zeidler's Universal Lexicon, 1732 (" a name applied to such a place in the sea as exhibits a heap of little islands lying close together, and almost hanging on to each other"). Atolu is probably connected with the Singhalese prep, elula, "inside" (Bell).
328-3 See Proc. Roy. Soc. Edin., 1879-80, No. 107.
329-1 [in table of Names of Atolls] The frequent termination du represents the Singh. duva, diva, and Sansk. dvipa, "island." Bell takes madulu for Sansk. mandala, "region." Qu. maha-atolu, " great atoll "?
331-1 The reseemblance to this is much closer to the old Singhalese with which it is compared in Mr Albert Gray's valuable paper already referred to.
331-2 "Usually"; but a Maldivian skipper who gave James Prinsep information wrote it from left to right (see Jour. As. Soc. Bengal, v. 794).
The above article was written by: Col. Henry Yule, C.B.