ESSAY ON THE VELOCITY OF LIGHT. 139 
even with telescopes of rather low power, which suffice to render the satellites 
of Jupiter very distinct when they are not eclipsed. We ought, however, to 
remark that it is never possible to observe the beginning and the end of the 
same eclipse of the first satellite, because, as we only see the umbral cone ob- 
liquely, a good portion of this cone, at the distance of the first satellite, is hidden 
from us by the planet itself. If we observe the satellite when it enters the 
umbral cone, it passes soon behind the planet, which prevents us from seeing it 
go out of the cone. If at other times we observe it when it comes out of the 
umbral cone, a short time before, and especially at the moment it enters the 
cone, it is hidden from us by the planet. When Jupiter is to the west of the 
sun, the umbral cone, which is always on the side away from the sun, is situate 
to the west of the planet; we therefore see the first satellite enter the umbral 
cone, but we do not see it go out, or, to use the language of astronomers, we 
see the zmmersion of the satellite, but we do not see the emersion. When 
Jupiter is to the east of the sun, it is the contrary that happens: the umbral 
cone is projected to the east of the planet, and we can observe the emersion 
whilst the immersion is invisible. 
Whilst in Padua, in 1610, Galileo discovered the satellites of Jupiter. The 
rapidity of the displacement of these little spheres to one side and to the other 
of the planet, the frequency of their extinction and of their reillumination, 
naturally excited the curiosity of astronomers, who accumulated observations 
on them. It was in studying these observations on the immersions and the 
emersions of the first satellite, in order to arrive at the means of predicting 
their eclipses, that Roémer made the capital discovery which we now propose 
to discuss. 
In order to fully comprehend the influence of the progressive transmission of 
light through space on the observations of the eclipses of the first satellite of 
Jupiter, and clearly to understand the manner in which Roémer was enabled 
to deduce the value of the velocity of light from the numerous observations of 
this phenomenon which had up to that period been tabulated, let us take an 
example relative to the progressive transmission of sound in our atmosphere. 
Suppose that the same sound is repeated regularly at equal intervals during 
any given time; this sound is produced, for example, by a wood-cutter who 
strikes with his axe a tree, or by a blacksmith who strikes with a hammer a 
piece of iron on his anvil. If we are quite near to the spot where the sound 
is produced we hear and see atthe same instant each blow. If we remove to 
a distance we no longer hear the blows at the same time we see them, and 
the interval of time which elapses between seeing and hearing the same blow 
becomes greater as the distance is increased. If, when far otf, we come near, 
then go off again, and then come near, and so on, the interval of time between 
seeing and hearing the same blow increases as we increase the distance, dimin- 
ishes as we diminish the distance, increases again with the distance, and so on: 
it thence results that, notwithstanding the regularity with which the blows of 
the axe or hammer are given, the successive sounds which we hear have not 
equal intervals, since the sounds which follow the blows which produce them 
are at intervals of time which alternately increase and diminish. When we 
walk towards the wood-cutter or the blacksmith, the sounds which we succes- 
sively hear are at shorter intervals of time than the blows of the axe or 
hammer, and the sounds are at greater intervals than the blows when we walk 
away from the wood-cutter. In walking with a regular step the circumference 
of a circle in the neighborhood of the place where the wood-cutter or the black- 
smith works, we will be sometimes further off and sometimes nearer to the place; 
and while we observe the blows succeed each other with great regularity, we 
hear the sounds occasioned by these blows succeed each other sometimes with 
diminishing intervals, and at other times with increasing intervals. ‘The irregu- 
larity in the succession of the sounds is rendered very sensible by the continuous 
