VOL. XLVIII.] PHILOSOPHICAL TRANSACTIONS. 393 



light. It is plain therefore, that on our hypothesis, any star must have a dif- 

 ferent apparent place for every different colour ; that is, its apparent disk must 

 be drawn out by the aberration into a longitudinal form, resembling the prismatic 

 spectrum, having its red extremity nearest its mean place. In the stars situated 

 about the pole of the ecliptic, its length should continue always the same, though 

 directed along all the different secondaries of the ecliptic in the course of a year : 

 but in those which lie in or near the plane of the ecliptic, it should be greatest at 

 the limits of the eastern and western aberrations, the star recovering its colour 

 and figure, when the true and mean places coincide. But there is no hope of 

 discovering, whether our system be true or false, by this consequence of it : for 

 the greatest length of the dilated disk being to the whole aberration, as the dif- 

 ference of the velocity of the red and violet to the mean velocity of light, i. e. but 

 about a 77th part of it, cannot much exceed the 4th part of a second. 



The time which the extreme violet light takes in arriving from any distance to 

 the eye, will be to that which the extreme red takes in coming from the same, 

 as 78 to TT . If Jupiter be supposed in a quadrate aspect with the sun, in which 

 position the eclipses of his satellites are most commodiously observed, his distance 

 from the earth being nearly equal to his distance from the sun, light takes about 

 41™ in passing from him to the earth ; therefore the last sensible violet-light, 

 which the satellite reflects before its total immersion into Jupiter's shadow, ought 

 to continue to affect the eye for a 77th part of 41*" ; that is, about 32» of time 

 after the last sensible red light is gone. It is therefore a certain consequence of 

 our hypothesis, that a satellite, seen from the earth, ought to change its colour 

 about half a minute before its total immersion, from white to a livid greenish 

 colour ; then into blue, and at last evanish in violet. It need hardly be ob- 

 served, that the same phenomenon must take place in the time of emersion by a 

 contrary succession of colours, beginning with red, and ending in white. 



If this phenomenon be perceived by astronomers, we shall have a direct proof 

 of the different velocities of the differently-coloured rays, and consequently a 

 mechanical account of their different degrees of refrangibility ; for he sees not, 

 to what other cause such an appearance could be reasonably ascribed. If it be 

 not, we may conclude, that rays of all colours are emitted from the luminous 

 body with one common velocity. 



Remarks on the Preceding Paper. By Mr. Short, p. 268. 



Ever since the above paper was delivered, Mr. S. had carefully attended the 

 emersions of Jupiter's first satellite through a reflecting telescope of 4 feet focal 

 length, and with a proper magnifying power ; but he had not [xerceived the least 

 alteration in the colour of the light reflected by the satellite, except in quantity. 

 It may indeetl be observed, that these emersions are seen sooner or later through 



VOL. X. 3 E 



