VOL. LXIII.] PHILOSOPHICAL TRANSACTIONS. ^ 420 



tion of the diaphragm, for determining the invisible segment. To repeat these 

 observations with judgment, it will be necessary to recollect the intention of 

 them ; which is, to measure what portion of the disc remains illumined, that is, 

 what portion of the satellite's light continues, though unperceived, to be trans- 

 mitted to the observer's eye, at the instant when the satellite disappears, on the 

 brink, of an eclipse. In lessening gradually the aperture of the glass, the obser- 

 ver should not begin with too small an opening; because the eye, not accustomed 

 to the great obscurity which follows, might not see the satellite at all. As the 

 opening is gradually contracted, the satellite seems to grow less. The observer 

 sometimes loses sight of it for a moment; but if he continues to look attentively, 

 he sees it again. The real disappearance is only to be concluded, when, on 

 fixing with steady eyes, for about half a minute, on the place it occupied, it is 

 seen no more; for if one persisted to observe it much longer, it might happen, 

 that it might be seen to glimmer at times, and immediately disappear. I have 

 always made it a rule, to consider the debilitation of the light, in this degree, as 

 actual disparition, and it is necessary that observers should agree on this point, 

 in order that their different estimations may be consistent. These fits of mo- 

 mentary glimmering and extinction are undoubtedly owing to the motion of the 

 particles of the atmosphere. In the clearest weather, there are always particles 

 of vapour floating in it, in vast abundance; according as these particles place 

 themselves in the direction of the ray of light, or out of it, the light of the sa- 

 tellite is diminished or restored, and the satellite, in consequence, is either hid 

 or rendered visible. This does not happen in eclipses, wherein a great part of 

 the light is in reality extinguished. But in the case I am now speaking of, 

 though the diminution of the aperture of the glass does indeed take away a great 

 quantity of it, yet this quantity is always relative to the actual state of the at- 

 mosphere : ( '' ) : if that state changes, this quantity becomes alternately sensible 

 or insensible, according as the light meets with more or less obstructions in its 

 passage from the vapours. Another thing, which it will be necessary to point 

 out, is, that the operation with the diaphragms, for determining the invisible 

 segment, must be made and concluded before the satellite has touched the 

 shadow. The proper time therefore for beginning this observation, will be de- 

 termined by the time the diameter takes in entering, which, in the perpendicu- 

 lar ingress, or when Jupiter is in the nodes, is 7™ for the first 2, and 1 1"' for the 



3d. 



The time of the oblique ingress is (7"") -j- for the first two, and (l 1™) "■ 



for the 3d ; which, for this last, may in extreme cases amount to about 27"" or 

 28™. It is proper to take 5"" more; for the observations of the diaphragm will 

 take up 2™, even when use has rendered them familiar, and the tables may be 

 2™ or 3™ behind. At present, it is sufficient to begin the observation id'" before 

 an eclipse of the 3d satellite; but there are times, in which it would be neces- 



