and rain, and from whence our seamen may have borrowed the expression; but it is evident that the order of derivation is here reversed, and that the phrase is taken from the name of the land in the neighbourhood of which such squalls prevail. In a Persian work of the year 1611 the name of Shamatrah occurs as one of those places where the Portuguese had established themselves; and in some very modern Malayan correspondence I find the word Samantara employed (along with another more usual, which will be hereafter mentioned) to designate this island.

PROBABLY DERIVED FROM THE SANSKRIT.

These, it is true, are not entirely free from the suspicion of having found their way to the Persians and Malays through the medium of European intercourse; but to a person who is conversant with the languages of the continent of India it must be obvious that the name, however written, bears a strong resemblance to words in the Sanskrit language: nor should this appear extraordinary when we consider (what is now fully admitted) that a large proportion of the Malayan is derived from that source, and that the names of many places in this and the neighbouring countries (such as Indrapura and Indragiri in Sumatra, Singapura at the extremity of the peninsula, and Sukapura and the mountain of Maha-meru in Java) are indisputably of Hindu origin. It is not my intention however to assign a precise etymology; but in order to show the general analogy to known Sanskrit terms it may be allowed to instance Samuder, the ancient name of the capital of the Carnatik, afterwards called Bider; Samudra-duta, which occurs in the Hetopadesa, as signifying the ambassador of the sea; the compound formed of su, good, and matra, measure; and more especially the word samantara, which implying a boundary, intermediate, or what lies between, might be thought to apply to the peculiar situation of an island intermediate between two oceans and two straits.

NOT ENTIRELY UNKNOWN TO THE NATIVES.

When on a former occasion it was asserted (and with too much confidence) that the name of Sumatra is unknown to the natives, who are ignorant of its being an island, and have no general name for it, the expression ought to have been confined to those natives with whom I had an opportunity of conversing, in the southern part of the west coast, where much genuineness of manners prevails, with little of the spirit of commercial enterprise or communication with other countries. But even in situations more favourable for acquiring knowledge I believe it will be found that the inhabitants of very large islands, and especially if surrounded by smaller ones, are accustomed to consider their own as terra firma, and to look to no other geographical distinction than that of the district or nation to which they belong. Accordingly we find that the more general names have commonly been given by foreigners, and, as the Arabians chose to call this island Al-rami or Lameri, so the Hindus appear to have named it Sumatra or Samantara.

MALAYAN NAMES FOR THE ISLAND.

Since that period however, having become much better acquainted with Malayan literature, and perused the writings of various parts of the peninsula and islands where the language is spoken and cultivated, I am enabled to say that Sumatra is well known amongst the eastern people and the better-informed of the natives themselves by the two names of Indalas and Pulo percha (or in the southern dialect Pritcho).

INDALAS.

Of the meaning or analogies of the former, which seems to have been applied to it chiefly by the neighbouring people of Java, I have not any conjecture, and only observe its resemblance (doubtless accidental) to the Arabian denomination of Spain or Andalusia. In one passage I find the Straits of Malacca termed the sea of Indalas, over which, we are gravely told, a bridge was thrown by Alexander the Great.

PERCHA.

The latter and more common name is from a Malayan word signifying fragments or tatters, and the application is whimsically explained by the condition of the sails of the vessel in which the island was circumnavigated for the first time; but it may with more plausibility be supposed to allude to the broken or intersected land for which the eastern coast is so remarkable. It will indeed be seen in the map that in the vicinity of what are called Rupat's Straits there is a particular place of this description named Pulo Percha, or the Broken Islands. As to the appellation of Pulo Ber-api, or Volcano Island, which has also occurred, it is too indefinite for a proper name in a region of the globe where the phenomenon is by no means rare or peculiar, and should rather be considered as a descriptive epithet.

MAGNITUDE.

In respect to magnitude, it ranks amongst the largest islands in the world; but its breadth throughout is determined with so little accuracy that any attempt to calculate its superficies must be liable to very considerable error. Like Great Britain it is broadest at the southern extremity, narrowing gradually to the north; and to this island it is perhaps in size more nearly allied than in shape.

MOUNTAINS.

A chain of mountains runs through its whole extent, the ranges being in many parts double and treble, but situated in general much nearer to the western than the opposite coast, being on the former seldom so much as twenty miles from the sea, whilst on the eastern side the extent of level country, in the broader part of the island, through which run the great rivers of Siak, Indragiri, Jambi, and Palembang, cannot be less than a hundred and fifty. The height of these mountains, though very great, is not sufficient to occasion their being covered with snow during any part of the year, as those in South America between the tropics are found to be. Mount Ophir,* or Gunong Pasaman, situated immediately under the equinoctial line, is supposed to be the highest visible from the sea, its summit being elevated thirteen thousand eight hundred and forty-two feet above that level; which is no more than two-thirds of the altitude the French astronomers have ascribed to the loftiest of the Andes, but somewhat exceeds that of the Peak of Tenerife.

(*Footnote. The following is the result of observations made by Mr. Robert Nairne of the height of Mount Ophir:

Height of the peak above the level of the sea, in feet: 13,842.
English miles: 2.6216.
Nautical miles: 2.26325.
Inland, nearly: 26 nautical miles.
Distance from Massang Point: 32 nautical miles.
Distance at sea before the peak is sunk under the horizon: 125 nautical miles.
Latitude of the peak: 0 degrees 6 minutes north.
A volcano mountain, south of Ophir, is short of that in height by: 1377 feet.
Inland, nearly 29 nautical miles.
In order to form a comparison I subjoin the height, as computed by mathematicians, of other mountains in different parts of the world:
Chimborazo, the highest of the Andes, 3220 toises or 20,633 English feet. Of this about 2400 feet from the summit are covered with eternal snow.
Carazon, ascended by the French astronomers: 15,800 English feet.
Peak of Tenerife. Feuille: 2270 toises or 13,265 feet.
Mount Blanc, Savoy. Sr. G. Shuckburgh: 15,662.
Mount Etna, Sr. G. Shuckburgh: 10,954.

Between these ridges of mountains are extensive plains, considerably elevated above the surface of the maritime lands, where the air is cool; and from this advantage they are esteemed the most eligible portion of the country, are consequently the best inhabited and the most cleared from woods, which elsewhere in general throughout Sumatra cover both hills and valleys with an eternal shade. Here too are found many large and beautiful lakes that extend at intervals through the heart of the country, and facilitate much the communication between the different parts, but their dimensions, situation, or direction, are very little known, though the natives make frequent mention of them in the accounts of their journeys. Those principally spoken of are: one of great extent but