PAST AND COMING TRANSITS AND ARCTIC EXPLORATIONS. 267 
he saw her whole circumference completed, by means of a vivid 
but narrow and ill-defined border of light, which illuminated 
that part of her circumference that was off the sun. He adds 
that it disappeared two or three minutes before the internal 
contact. A similar phenomenon was witnessed during the same 
transit by Wales and Dymond at Hudson’s Bay, by Pingre 
and De Fleurien at Cape Francis, in the Island of St. Domingo, 
and by various other observers at different places.” 
A little consideration will suggest the true cause of this 
appearance, and will show that its effect on the observation of 
internal contact cannot but seriously affect the accuracy of the 
timing. No light can show itself round the portion of Venus out- 
side the sun, between the moments of exterior and interior con- 
tact, unless the planet has an atmosphere capable of refracting 
the solar rays ; but if the planet has such an atmosphere, the 
observed effect cannot but be produced. If we suppose an 
observer on Venus at a point, P, on the part of her limb most 
remote from sun — P not being the point which is seen from the 
earth at that part of the limb, but so placed that the true 
horizon-plane for an inhabitant of Venus there is parallel to 
the line from the observer on earth to the centre of Venus— «- 
then, if there were no atmosphere, an observer at P could see 
neither the sun nor the earth, at least not where the terrestrial 
observer is placed. The sun would be just below the true 
horizon of the observer on Venus, and so would the observer on 
earth ; and a fortiori the observer on earth would not be able 
to see the sun round the part P of Venus. This part would be, 
as it were, the summit of a hill, from which the sun on one 
side and the earth on the other would be invisible, and therefore 
invisible from each other. But if there is an atmosphere on 
Venus resembling our own atmosphere in its effects, the observer 
at P would see the sun raised by refraction above his horizon, 
and the earth directly opposite raised considerably above the 
horizon — precisely as at the time of total lunar eclipse the 
observer on earth, if so placed that the eclipsed moon is appa- 
rently just above the horizon (really raised by refraction), can 
see the sun also directly opposite the moon, and raised wholly 
above the horizon. And as the line of sight from the observer 
at P to the sun on one side, and to the earth on the other, is 
thus twice bent by refraction, so the line of sight from the 
observer on earth to P is curved doubly as it passes P, and the 
sun is brought into that observer’s range of view. The same is 
true for all points round the part of Venus’s limb outside the 
sun, the atmosphere of Venus bringing greater and greater 
quantities of sunlight round the dark limb the nearer the part 
of the limb is to the sun. Thus there is seen round the arc of 
