Royal Astroiwmical Society. 473 



pearance of a comet at the time of the death of Conatantine Ducas, 

 in May 1067. We have no particulars, and even the reality of the 

 comet is subject to some doubt. In 1066, a year memorable in 

 English history as that of the Norman Conquest, a very grand and 

 remarkable comet is recorded by nearly every chronicler and historian 

 of the age. It was observed throughout Europe and China during 

 the months of April and May. The elements of Halley's comet, as 

 they at present exist, will hardly represent these circumstances with 

 a sufficient degree of accuracy ; but if we assume the following 

 numbers, we shall have a very fair agreement between observation 

 and calculation : — 

 Perihelion passage 1066, April l** 0**, Greenwich time, Julian style. 



o / 



Longitude of perihelion 264 55 



Ascending node 25 50 



Inclination 17 



Least distance 72 



Motion retrograde. 



In this orbit the longitude of perihelion is further in advance than 

 that of the present orbit of Halley's comet by 30°, and the node by 1 7° : 

 the perihelion distance is greater by 0"14. But are these differences 

 to be considered beyond the limits of probability ? In the lapse of 

 so many centuries, may not the planetary perturbations have pro- 

 duced alterations in the elements at least equivalent to those here 

 exhibited, especially since there is good reason to conclude that the 

 plane of the orbit of Halley's comet formerly coincided much more 

 closely with the plane of the ecliptic than at present, the comet 

 being therefore subject to far larger perturbations than it has under- 

 gone in more recent times? 



The elements which I have assumed for the comet of 1066 have 

 been obtained partly from the observations in that year, and partly 

 from the Chinese description of a comet a.d.141, which also agrees 

 with the supposed period of Halley's. The same orbit will represent 

 the apparent paths in both years ; and I would particularly insist on 

 this point as one of some importance in the present inquiry. I may 

 also remark, that it appears by no means improbable that an orbit 

 more closely resembling that of Halley's comet might be made to 

 represent the circumstances recorded of the comet of 1066 with, 

 tolerable accuracy. The descriptions of this object are so confused, 

 that a good deal of uncertainty is necessarily attached to any con- 

 clusion we may deduce from the observations. 



The preceding return of Halley's comet took place, I think, in 

 989, and Burckhardt's calculations relative to the comet observed in 

 China in that year strongly support this idea. The perihelion 

 passage would occur about the 1 2th of September. 



In 912 a comet was seen in China, in the month of May, in Leo, 

 near the star marked )( on our charts. It was also perceived in 

 Europe. The orbit of Halley's comet, with very trifling alterations, 

 will agree with the above position about May 13 or 14, if we fix the 

 perihelion at the beginning of April. 



PhiL Mag. S. 3. Vol. 36. No. 245. Jime 1850. 2 I 



