the Scintillation of the Stars. 221 
3rd. Selon to be avoided. 
It often happens that from one night to another the scintil- 
lation varies very considerably ; but it augments or diminishes 
proportionably for all the stars, except perhaps for those which, 
being nearest the horizon, have always a strong scintillation, or 
except when accidental causes modify it momentarily. Among 
these accidental causes we may cite, first, the twilight, which 
almost always very much augments the scintillation; and 
secondly, the neighbourhood of clouds. I think M. Kaémtz 
was the first to notice that the scintillation augments when there 
are clouds driven by the wind. This is the fact, as I have ob- 
served it in thousands of cases, and I do not remember noticing 
a single exception. 
Hence I do not say that we ought to reject observations taken 
when the stars are near clouds ; only we must bear in mind this 
circumstance, as it greatly modifies the results obtained. A 
bright moonlight is also very unfavourable; for observations 
taken when the moon is full, are much less exact than those 
taken when she is absent. 
4th. Comparison of the observations made by different persons. 
This is the delicate point to consider ; for what precedes suf- 
fices for the study of the phenomenon when all the observations 
are taken bya single person ; but when there are several, how are 
we to know if the scintillation designated 2°5 is equal to that 
of another person also designated by 2-5? I believe it is im- 
possible to obtain this unison; and unless every observer could 
be taught by one single person experienced in the matter, I am 
certain it never will be obtained. However, here is a method of 
recognizing if the scintillation is, in absolute value, stronger in 
one station than in another :— 
At Morges, during the nights of maximum scintillation, the 
stars at the zenith scintillate very decidedly. In the nights of 
middling scintillation the stars in the same position scintillate 
feebly, though always enough to be appreciable; but in the 
nights when the scintillation is at its minimum, the stars 
nearest the zenith have no longer any scintillation at all; and 
the nearer the scintillation approaches to its mmimum, the more 
extended is the spherical segment (of which the zenith is always 
the centre), which comprehends all those stars of which the 
scintillation is inappreciable. I have sometimes seen, when the 
scintillation was very feeble, that the stars lost all scintillation 
as soon as they were at 42° above the horizon; but I have never 
seen it cease entirely for stars less elevated, though, from what 
Arago says, it appears that that happens sometimes. 
He names, among others, the observations of M. de Hum- 
