I 
464 NEW OBSERVATIONS OF THE PLANET MERCUET. 
of a planet whose phase was constantly increasing, a halting or even retroo^ression in 
that phase about that time. 
Now in Table I. we see that t + m became 90° between February 16 and 17. 
Turning to Table VI. we note that between February 16 and 17 occurred a sudden 
halt and retrogression in the phase, 
48. Phase of Measures. — Thirdly, it will be noticed that the micrometric 
measures do not keep step with this halting of the measures got from the drawings. 
The former keep on more evenly, are bigger than the latter up to February, and 
then lose. This is precisely what should happen if the planet were an ellipsoid 
with its longest axis at the point of mean libration; for this was all the time getting 
squarer to the line of sight up to February, in consequence of which the equatorial 
diameter would increase up . to that time, thus offsetting the loss on the terminator 
le to that terminator's passing through the minor axis of the ellipse, which fact 
d 
taken account of hy the drawing 
49. Mercury 
All this is interesting from being app 
ently but a curious concatenation of circumstances ; for though there is very Httle 
doubt that Mercury is such an ellipsoid, its longest axis pointing to the Sun, its 
middle one being perpendicular 
plane of the ecliptic, and its shortest 
being the planet's polar axis, there is also but little doubt from theory that the 
difference in the diameters is much too small to be visible, even indirectly. For 
each of these phenomena may be otherwise explained. ■ Article 46 is explicable by 
greater irradiation at the centre of the limb; Articles 47 and 48, by the passing 
of Testudo regie across the terminator. 
50. Summary. — These observations, therefore, go to show that Mercury is : 
(1) A body somewhat larger than has been supposed ; its polar diameter sub- 
tending at distance unity about 7".50, e<juivalent to about 3,400 miles. From the 
difficulty of evaluating the irradiation I consider that this is probably too large, and 
that 3,300 miloB is nearer the truth. The older measures, however, are certainly 
too small. 
<^ 
(2) Of a probable density much less than the Earth's, and according, within errc 
of observation, with what a ^nW principles of size and solar position would lead ns 
to suppose ; 
Of a probable mass, from (1) and (2), between the older and later determina- 
tions, — about 2V of the Earth's : 
(4) Rotating in 87.969 days, its orbital period- The decimal is here given 
9 
O 
because though the period of rotation may differ ty a minute amount from the 
