F’. A. P. Barnard on the Zodiacal Light. 227 
isnot remembered. Now had the light been at that hour but 
barely perceptible in the horizon, it is manifest that the intersec- 
tion of the tangent limiting the earth’s shadow with the outer 
stratum of the ring, could not have been below the horizon. 
Supposing then that intersection to have been situated exactly at 
the horizon, its distance from the earth’s centre would be equal, 
in miles, to 39564/2, or 5596 i—giving a distance above the 
earth’s surface of 1640 miles. Supposing the angular altitude 
' of the light at the same time to have been thirty degrees, the dis- 
tance of the apparent summit from the earth’s surface must have 
been 3434 miles, or nearly equal toa radius of the earth; and al- 
lowing it to have been forty-five degrees, the same distance would 
have been 4890 miles. 
The singular discrepancies to which the different observations 
Peonduct us, lead us almost irresistibly to the conclusion that the 
ting hypothesis is untenable. They cannot be explained but by 
making assumptions in regard to the reflecting power of the nebu- 
: losity, such as to destroy the value of all observations upon this 
phenomenon, and to conflict, at the same time, with established 
' laws of optics 
| 
But, in the fourth place, in accepting the hypothesis of a ring, 
we must accept along with it, not merely those consequences 
Which it may be convenient to us to admit, but all the conse- 
quences which it legitimately involves. We must show that the 
visible arch which we regard asa portion of this supposed ring, 
conforms, under all circumstances, to the geometrical conditions 
Which the hypothesis imposes. ‘That we may be able to judge 
ow far this is the case, let us consider how these conditions may 
be investigated. 
of the ring about the line which joins the centres of the earth 
and sun, and which, being produced, forms the axis t 
Shadow. In this spherical surface the ring must always be found, 
cylindrical) with the same 
spherical surface, will be a small circle of the sphere ; and the 
i eee 
we 
wy, 
eb8erver, and it will be at once self- 
evident that the arc-radius of one of 
small circles will be comple- 
mentary to that of the other. If, in 
~ figure, HR be the projection of 
circle of the horizon and RT, 
~ 
