and afterwards the comet would rise in daylight, and could not therefore be observed.

The positions of the comet observed by Mr. Pogson do not well accord with the calculated places of either part of Biela's comet, or of the meteor-stream through which the earth passed on November 27. Capt. Tupman (R.A.S. Notices, xxxiii. p. 318, March 1873) gives reason for his opinion, that the body seen by Mr. Pogson was neither Biela's comet nor a meteoric aggregation travelling in the same orbit, nor a body that had passed near the earth on November 27. Dr. Oppolzer (Ast. Nach., Nos. 1920 and 1938, January 31 and May 13, 1873), although he originally held the same view, was led by the investigations he undertook to consider it highly probable that Pogson's comet is closely connected with the shower of shooting-stars on November 27, and that it is even possible that it was a head of Biela's comet; but Dr. Bruhns (Ast. Nach., No. 2054, September 10, 1875) arrives at the conclusion that it is very probable that Pogson's comet was unconnected with Biela's comet or with the meteor-swarm, and that it was a new comet.' If it were Biela's comet, the latter must have been about twelve weeks behind its time. It has been suggested that the observations made on December 2 and 3 referred to different heads of the comet, but there seems no doubt that the body observed was the same on both occasions. In any case, the fact that

Professor Klinkerfues should have felt sufficient confidence in the truth of his hypothesis to send the telegram to Mr. Pogson, and that the latter should have actually detected a comet in the neighbourhood of the position indicated, forms a very striking, I might almost say, romantic episode in astronomical history, whether the body thus found was a portion of Biela's comet, or a meteor-swarm on its track, or even an independent body.

In consequence of the interest excited by the above observations, Mr. Hind communicated to the Royal Astronomical Society (Notices, xxxiii. p. 320) an account of the actual state of the calculations with regard to Biela's comet, from which it appears that 'both nuclei of the comet were last observed in the autumn of 1852, having been found much further from their calculated places than was expected, a circumference which undoubtedly affected the number of observations, and which was occasioned by the unfortunate substitution by Professor Santini of a semi-axis major depending wholly upon the observations of the previous appearance in 1846, in place of that which he had deduced from observation in 1832, and carried forward by perturbation to 1846. This source of error in the prediction for 1852 is indicated by Professor Santini in a communication made to the Venetian Institute in November 1854. There is no reason to suppose that any perturbations beyond those resulting from known causes operated between the appearances of the comet in 1846 and 1852; indeed, the observations of these years have been connected without difficulty by the application of planetary perturbations during the interval.' The effect of the perturbations was calculated by Professors Santini, Clausen, Hubbard, and Michez, for the period from 1852 to 1866, and the perihelion passages were

fixed for May 24, 1859, and January 26, 1866. In 1859,' Mr. Hind proceeds, 'the position of the comet in the heavens rendered its discovery almost hopeless, and its having passed by us unobserved is thus accounted for; but it is not so as regards the return in 1866. I believe it is certain that the comet did not pass its perihelion in that year within seven weeks of the time predicted... So far as I know at present the calculation of perturbations from 1866 to 1872 has not been undertaken by anyone... it has probably been felt to be a useless labour to carry forward the elements from the predicted time of perihelion in 1866, considering the want of success attending the endeavours to find the comet in the corresponding track.... If we suppose that the comet did really encounter the earth [on November 27, 1872] in descending to perihelion on December 27, there will be found since 1852 three mean revolutions of 6.754 years, and the perturbations being small from 1866 to 1872, the comet might have been in perihelion about March 28, 1866, instead of January 26. It is clear, therefore, that if the perihelion passage of Biela's comet took place in 1866, six or eight weeks later than anticipated, its having passed unobserved need occasion no surprise.'-ED.]

SECTION IV.

DOUBLE COMETS MENTIONED IN HISTORY.

Is there any example in history of the division of a comet into several parts?—The comet of B.C. 371-Ephorus, Seneca and Pingré-Similar observations in Europe and China-The Olinda double comet, observed in Brazil, in 1860, by M. Liais.

THE doubling of Biela's comet did not fail to direct attention to the several instances on record of analogous phenomena which had hitherto been looked upon as little worthy of belief. It was then remembered that Democritus had, according to Aristotle, related the fact of a comet having suddenly divided into a great number of little stars. It was this, perhaps, that gave rise to the opinion of certain philosophers of antiquity that comets were composed of two or more wandering stars. Seneca, in endeavouring to refute this opinion, mentions the account given by Ephorus, the Greek historian, of the division of the comet of the year B.C. 371 into two stars. He thus expresses himself:—

Ephorus, who is far from being an historian of unimpeachable veracity, is often deceived-often a deceiver. This comet, for example, upon which all eyes were so intently fixed on account of the immense catastrophe produced by its apparition the submersion of the towns of Helice and BuraEphorus pretends divided into two stars. No one but himself has related this fact. Who could possibly have observed at

what moment the comet dissolved and divided into two? And besides, if this division was actually seen to take place, how is it that no one saw the comet form itself into two stars? Why has not Ephorus given the names of these two stars?'

These two last arguments appear of little value, whilst the fact itself mentioned by Ephorus, since the observations of January 1846, appears no longer impossible.

In 1618 several comets were observed, some in Europe, others in Persia, which could not be identified or distinguished from one another with certainty. Two of these comets were seen at the same time and in the same region of the heavens, and this led Kepler to suspect they were parts of one and the same comet, which had divided into two. When recording this opinion of Kepler's, Pingré, who took part with Seneca against Ephorus, now considers the great astronomer at fault, and exclaims, 'Quandoque bonus dormitat Homerus!' At the present time, although unable to affirm that this division did actually take place, we are forced to consider the conjecture of Kepler as at least probable. There are, besides, other facts on record not altogether dissimilar, and which are narrated by Pingré himself :

'In the year B.C. 14, Hantching-Ti ascended the throne of China, in the twenty-sixth year of the fourth cycle; in the eighteenth year of his reign a star was seen to resolve itself into fine rain and entirely disappear.'

Under the consulate of M. Valerius Massala Barbatus, and P. Sulpicius Quirinus, before the death of Agrippa, a comet was seen for several days suspended over the City of Rome; it then appeared to resolve into a number of little torches.' This is related by Dion Cassius.

According to observations recorded by the Chinese annalists, and collected by Edward Biot, three comets joined together appeared in the year 896, and described their orbits.

in company. Here again is a passage from Nangis, quoted by Pingré, from which it appears that the comet of 1348 separated into several fragments: In the early part of the night, in our presence and to our great astonishment, this very large star divided into several beams, which spread eastward over Paris and entirely disappeared. Was this phenomenon a comet, or other star, or was it formed of exhalations, and again resolved into vapour? These are

questions I must leave to the judgment of astronomers.'

It is very probable that some of the phenomena of sudden

division mentioned by the ancients may have reference to bolides or, as Pingré says, to meteors'; but the facts are none the less curious, since between comets, bolides, and shooting stars a real relation and community of origin have been proved to exist.

But, before closing this section, we must not forget to mention a phenomenon analogous, if not to the doubling of Biela's comet, at least to the fact of two or more nuclei existing in the same comet. The observations were made at Olinda in Brazil by M. Liais, a contemporary French astronomer, in the