M. Secchi on Shooting-Stars, 379 



perpendicular ought to remain the same when the elevations of 

 the two stations are employed ; the degree of accordance between 

 the two results serves as a criterion of the trustworthiness of the 

 observation. Repeating the same construction for the end of 

 the trajectory, another point is ascertained, after which the 

 direction and inclination of the actual trajectory are readily de- 

 duced. 



In order to determine the coordinates of altitude and azimuth, 

 the trajectories were first transferred to a celestial globe 0*53 

 metre in diameter, the apparent trajectories corresponding to the 

 two stations being marked with different colours. Afterwards 

 all the altitudes and azimuths, of the sixty-nine contemporaneous 

 trajectories which were visible from beginning to end, were 

 determined on the globe itself (placed according to the sidereal 

 hour, which had previously been calculated from the mean times 

 of observation) by means of a moveable vertical quarter of a 

 circle. 



These preliminaries completed, I proceeded to the gra- 

 phical constructions, referring all to the horizon of Rome, con- 

 sidered as parallel to that of Civita Vecchia ; for the difference 

 scarcely exceeds half a degree, the distance between the stations 

 being merely 65 kilometres. 



The results of these constructions showed that in a great 

 number of cases the values of the perpendicular heights agreed 

 tolerably well, or at all events that they could be made to agree 

 by supposing very probable errors of observation amounting to 

 1 or 2 degrees. There were many cases, however, in which this 

 agreement was inadmissible. Nevertheless, as no doubt could 

 be entertained of the contemporaneity of the observations, and 

 as the general directions of the apparent trajectories accorded with 

 those which parallax required, we were compelled to admit that 

 the same point of the actual trajectory had not in reality been 

 observed, but that the star had been seen from the two stations 

 at different points of its course. 



This conclusion involves no improbability ; for, first, the light 

 of many stars is very feeble at the beginning arid end of its 

 visible course, and a difference of distance amounting to 60 or 

 70 kilometres might easily render it invisible from one of the 

 two stations; secondly, notwithstanding all possible care and 

 attention, the eye perceives the star only after its course has 

 commenced, and it is not at all uncommon to find disaccordance 

 even between two observers at the same place ; and lastly, the 

 strength of the observer's vision must affect essentially the 

 result. This is proved indirectly by the fact that, for the 

 observed ignition, at the middle of its course, of a star which 

 burst into a red-coloured flame and was seen from both stations, 



2C 2 



