24 REPORT — 1861. 



of the solar eclipses. He relied on the two eclipses of the je&t — 719. In that of 

 March 8, it is recorded that the middle of the eclipse took place at Babylonian 

 midnight. As we know from the astronomical tablets found oy Mr. Layard that 

 the Babylonians measured the time from apparent noon by clepsydras, which ran 

 out in two hours, the time of midnight, being at the end of the sixth kazb, or 

 running out of the clepsydra, would be known with gi-eat precision, so that an 

 error of above a few minutes in the time of the middle of the eclipse is inadmissible. 

 An error of twelve minutes can scarcely be supposed possible. This, however, would 

 represent an errorof about325"in the mean elongation,or about0"-57in the coefficient 

 of^the acceleration ; while no error which we can suppose possible in the motion of 

 the node would sensibly affect the result. It appears from this, that the record of 

 this eclipse enables us to determine the quantity of the acceleration with far more 

 accuracy than do any of the records of solar eclipses. Dr. Hincks has calculated 

 the circumstances of this eclipse from Hansen's tables ; and without laying much 

 stress on his calculations, which required to be verified, he would state that the 

 middle of the eclipse took place about 11 p.m. Babylonian time, and that the 

 moon must have been completely out of the shadow before midnight. The other 

 eclipse was on the 1st of September ; and it is recorded that it began after the moon 

 had risen. This statement is one about which a mistake would be impossible. 

 And yet, according to the author's calculation from Hansen's tables, the moon was 

 more than two digits eclipsed when she rose at Babylon. But this is not all. He 

 argued that Ptolemy's reasoning respecting this eclipse implied that his records 

 stated that the eclipse did not commence till some time, probably half an hour, 

 after the moon had risen. The sun, he says, set at seven ; consequently the eclipse 

 began at half-past seven, Babylonian time. If then the author's calculation be 

 correct, there is an error of more than an hour in the time when this eclipse 

 commenced, which of course must be occasioned by Hansen's having overrated the 

 acceleration. The author concluded by saying that he did not expect or wish that 

 his calculations should be held conclusive ; but he wished that others should make 

 the calculations, and that these important data should not be ignored, aa they had 

 hitherto been. 



Cases of Planetary Instability indicated by the appearance of Temporary 

 Stars. By Daniel Yaughajt. 



In a paper which I sent to the British Association foitr years ago, I ascribed the 

 perpetmty of the sun's light to the combustion of ether collected fi-om space by his 

 attraction, and compressed at his sirrface to a density sufficiently great tor the play 

 of chemical forces. I showed that the brilliancy of liuninous meteors is the neces- 

 sary consequence of the great pressui-e which rapidly moving bodies impart to the 

 envelope of this luciferous fluid belonging to the earth, and that the dormant pho- 

 tospheres which it forms for dark central bodies, when traversed by worlds in the 

 last stage of existence, must give birth to a vast iUmnination, exhibiting to us the 

 pecidiar characters of the temporary stars. The main results of my subsequent re- 

 searches on the subject have been published in the ' London, Edinbm-gh, and Dublin 

 Philosophical Magazine ; ' and in two communications which appeared in the Num- 

 bers for December 1860, and April 1861, the more obvious cases of the instability of 

 satellites fi-om the reduced size of their orbits have been investigated, partly with 

 a view of tracing the com-se of great meteoric displays in the heavens. In these 

 I have supposed the central orb vastly superior to its attendant in mass and dimen- 

 sions ; but on examining the consequences which must ensue when the dispropor- 

 tion between both bodies is not so gi-eat, we render more satisfactoiy the explana- 

 tion of temporally stars, without exceeding the limits of reasonable probability in 

 estimating the number of dai-k systems in our imiverse. 



In these investigations the hypothesis of solidity, in the sense usually received, 

 should be abandoned. In very close proximity to a great central sphere, a satellite 

 as large as the earth could not be considered imyielding, though composed of the 

 strongest materials known ; nor could it deviate much fi-om the foi-m which it might 

 assume if reduced to a state of fluidity. In a very small orbit also, the disturbing 

 forces operating on the yielding mass would have the ultimate effect of bringing 

 the same point of its surface into pei-petual conjunction with the primary planet. 



