count of the refraction of light in the atmosphere; and this contributed materially to give them a superiority in point of accuracy over those of the Greeks. It is true, that the phenomena of refraction must have been known to Ptolemy, since he speaks of them in his work on Optics, but he does not notice the subject in the Syntaxis where the application of it would have been of great importance; and this has induced an opinion that he had not discovered those phenomena till after the composition of that work: it is evident, therefore, that the observations made by this astronomer, as well as by all those who preceded him, on the apparent places of stars, must be affected by this cause of error. It is worthy of observation also, that some of the phenomena of the refraction of light must have been known before the time of Ptolemy; since Cleomedes makes an allusion to the wellknown experiment of rendering visible a ring, or other object, when placed in a vessel, by filling the latter with water: this is the first notice extant concerning the effects of refraction; but it is connected with the erroneous notion, prevalent among the ancients, that in vision, the rays of light proceed from the eye to the object. Alhazen, the author of a treatise on refraction, was the first writer who gave a reason for the apparent magnitude of celestial bodies in the horizon; he ascribes it to a tacit judgment formed by the mind from the idea which, in that situation, is entertained of great distance; and this, he observes, is excited by the number of terrestrial objects then perceived between the celestial bodies and the spectator.

a

The astronomy of the Arabians was, at the decline of the empire of the Khaliphs, and about the time of the revival of learning in Europe, cultivated in the northern part of Persia: from an Arabian MS., the writer of which lived in the thirteenth century, we learn that a grandson of Gengis-khan took into his service Nassir-Edin, an Arabian or Syrian astronomer, by whose direction an observatory was built at Maragah where, from the year 1261, were made a number of celestial observations which the mathematicians of the prince used for correcting the tables of Ptolemy; these new tables were afterwards published and, being dedicated to the Tartarian Prince Ilchan, they, from him, received the denomination of the Ilcanic tables. The observatory itself De Mundo, Lib. II. cap. 5.

was a remarkable structure, since it formed a great sun-dial, being crowned with a dome, having an opening through which the rays of the sun passing, indicated the hour of the day and the altitude of the luminary. The instruments used in the observations appear to have been similar to those of Ptolemy and the Arabians; the writer of the MS., in enumerating them, mentions an armillary sphere which was to be employed when the celestial body was not on the meridian, and on whose alidade he proposes that there should be a tube to protect the eyes; probably meaning that it should contain some coloured transparent substance similar to the dark glasses at present employed when viewing the sun. In describing an instrument to measure the diameter of the moon, he says its alidade should carry two plates, of which that next to the eye should have a small aperture pierced in it; the other, a larger, and the latter plate should be moved forward or backward till the full moon exactly filled its aperture; then a scale graduated on the alidade served to determine the value of her diameter.

The last circumstance of importance connected with the astronomy of that part of the world is the construction of a magnificent observatory at Samarcand, about the year 1400, by Ulugh Bey, a descendant of Tamerlane: this prince caused the obliquity of the ecliptic to be measured with a gnomon of great dimensions, and a catalogue of stars to be formed; in the preface to this work, it is stated that there are eight stars marked in the catalogue of Ptolemy which could not then be found in the heavens, and among these are mentioned six unformed stars near the southern fish. Bailly observes that these six have not since been marked in any catalogue; whence he thinks it probable that they may have disappeared between the times of Ptolemy and Ulugh Bey, for, since four of them were, according to the former astronomer, of the third magnitude, it is not likely that they would have found a place in his catalogue if they had not then been visible in the heavens. A work on astronomy, composed by order of the Tartarean prince, was, in 1665, translated by Greaves and Hyde: it contains a series of tables which were considered preferable to those of Nassir-Edin, and are admitted by Delambre to be tolerably exact for the age in which they were computed.

CHAPTER XV.

THE ASTRONOMY OF THE ANCIENT HINDUS.

The antiquity of the Hindu astronomy inferred from their tables.-The four Hindu ages supposed to be founded on astronomical periods.-Probable source of the eastern fables concerning the renewals of the earth.-Situation of the point at which the Hindu zodiac commences.—Length assigned by the Hindus to the solar year.-Epochs of Hindu tables.-Erroneous value assigned by the Hindus to the precession.- Observations on the elements of the Hindu Tables.-The planetary system of the Hindus similar to that of the Greeks.-Instrumental observations were made by the Hindus. - Lunar periods known to this people.-Probability that the Hindu astronomy is formed by improvements on that of the Chaldeans, Greeks and Arabians.

THE great antiquity of the Hindu astronomy has been the subject of much discussion among the learned of Europe; on the one hand, the evidence afforded by the tables brought from the countries on this side of the Ganges leads to an opinion that the science had been cultivated there long before the Noachian deluge is supposed to have taken place; and on the other hand, the entire absence of all notice of these tables and, indeed, of any circumstance connected with the science or history of a period so remote, not only in any Sanscrit writing known to exist, but in any document relating to the learning of the neighbouring countries, Assyria and Persia, where the practice of observing the heavens, in very ancient times, was diligently pursued, and where such works, if then in being, could not have been unknown, gives rise to a well-founded suspicion that the tables may have been fabricated in a later age. We are to add to this, and the fact is highly confirmative of the suspicion, that in the immense interval between the epoch of the tables and the probable origin of the modern astronomy in India, no traces appear that the science was cultivated in that part of the world, where yet it is supposed to have been, previously, brought to a state of great perfection. Even if we admit all that has been alleged concerning the antiquity of astronomy; since every monument of its

existence must have perished when the first philosophers from Europe visited the East, that science was, to them, as though it had never been; it was to be anew discovered, and its gradual rise can only be traced in the few accounts which have been transmitted to us, of the Egyptians, Chaldeans and Greeks.

But if the astronomy of the Hindus pretends to an antiquity. which must be allowed to be of a doubtful character, their chronology, if understood literally, goes back beyond all the limits of probability. In the Institutes of Menu, the most ancient Sanscrit work in existence, and which Sir William Jones supposes to have been written 880 years Before Christ, is a passage from which we learn that this people supposed there had been, and would continue to be, an innumerable succession of periods of time, each consisting of many millions of years, and each terminated by a destruction and a re-formation of the universe. These periods are divided into yugas or ages of various lengths, but all of them are multiples of the least, which is called the Cali-yuga, and is said to be equal to 1200 years of the gods, each of which the Hindus pretend to be equal to 360 years of mankind; one of the latter years being supposed to be a day of the gods, and 360 such days constituting one of their years; according to this mode of reckoning, therefore, the Cali-yuga should consist of 432000 common years. Ten of these periods constitute one Maha-yuga, and one hundred or one thousand Maha-yugas, for the number is variously stated, make the period called calpa, which is said to be one day of Brahma. The Hindus suppose the present world to have been created at the commencement of this last period, and nearly half the period to be now past. The Maha-yuga is divided into four parts, of which the most ancient, called Satya-yuga, or the golden age, consists of four Cali-yugas; the next, called Trétá-yuga, or the silver age, consists of three; the third, called Dwápar-yuga, or the brazen age, of two; and the fourth is the Cali-yuga, in which we live; this is denominated the iron, or corrupt age; and these denominations afford another proof in addition to the many that may be offered of the propensity of men to consider the world as having deteriorated in the course of time; and their own, as worse than any of the preceding ages.

All Europeans who have commented upon these Hindu periods agree in considering them as astronomical, but they differ much in their manner of interpreting them; an imperfect knowledge of the retrogradation of the equinoctial points, which the Hindus, like the Arabians, supposed to be at the rate of 54 seconds yearly, might have led to an opinion that those points performed a revolution through the ecliptic in 24,000 years; and this period, multiplied by 18, which is the number of years expressing nearly the time of a revolution of the moon's nodes, would have given 432,000 years, the value of the fundamental cycle above mentioned: and Mr. Davis, in a dissertation on the astronomical calculations of the Hindus a, argues, from passages in the Sourya Siddhanta, that the Maha-yuga is a period formed by the continual product of the times expressing the periodical revolutions of the equinoxes, of the moon's nodes and apogeum: he also considers the Calpa to express, in like manner, the time to elapse between two consecutive conjunctions of all the planets with their nodes and apogea, at the commencement of the zodiacal constellations. The values, moreover, which have been above assigned to the Hindu cycles do not appear to have been universally adopted in India; for, according to Mr. Crawford", the Hindus of Bali estimate the Maha-yuga at 15,025 years, instead of 4,320,000 years; and this period has, very probably, been formed from some erroneous determination respecting the movement of the equinoxes.

But though we should admit that the values of the above periods may have been given for those of certain astronomical cycles, it is no less probable that they have some relation to circumstances in the political history of the Hindus; the division of time into four ages is not peculiar to the people of India; indeed it was universally adopted by the ancients, and the periods in all cases seem to have designated the durations of particular races of sovereigns. Thus, in the fabulous chronology of Egypt given by Manetho, we find mention made of three ages, preceding those of the historical times, during which the throne of the country was said to have been occupied by gods, demi-gods and Manes: from the Zend-Avesta we learn that the ancient Persians Asiatic Researches, vol. ii. b Ibid. vol. xiii.