292 Variation in the Obliquity of the Ecliptic. [July, 
centre of gravity be not coincident with the centre of the 
sphere, eccentric motion occurs. 
In the conclusions arrived at relative to the movement of 
the earth’s axis being such as to produce only one uniform 
course in the axis, we fail to notice any reference, or any 
consideration, to the fact that the centre of gravity of the 
earth does not lie in the plane of the Equator. 
But, again, let us examine the globe, and taking half the 
northern hemisphere—viz., from longitude 15° W. round to 
about 165° E.—we have the mass of Europe, Asia, and 
Northern Africa on one side of the sphere, whilst on the 
opposite side we have only the continent of North America. 
About three times as much land is above the ocean between 
15° W. longitude and 165° E. longitude as there is in the 
other half, and it therefore follows that the earth’s axis 
cannot pass through the centre of gravity of the earth. 
We have, then, at present a spheroid of irregular form, 
in which the centre of gravity does not lie in the plane of 
the Equator, nor does the axis of rotation pass through the 
centre of gravity. 
Let us now ask a question. There are forces acting by 
attraCtion externally on the earth, and supposed to produce 
a conical movement of the axis, in consequence of the pro- 
tuberance at the Equator. Will the movement of the axis 
of diurnal rotation be exactly the same, no matter whether 
the centre of gravity of the earth lies in the plane of the 
Equator, and is passed through by the axis of rotation, or if 
this centre of gravity is located elsewhere? We need not 
remind the reader that the movement would not be 
the same. 
Of all the conditions that produce a change in the second 
rotation of a rotating body—that is, in the change in direc- 
tion of the main axis of rotation—none is greater than that 
of a displacement of the centre of gravity, or the fact that 
the centre of gravity of the rotating body is not coincident 
with the centre of the sphere. The divergence from a direct 
course or vertical plane of a spherical shot is always great 
when the centre of gravity of this shot is not equidistant 
from all parts of the surface, and this divergence is accom- 
panied by—if not mainly dependent on—a rotation of the ° 
shot, in a manner dependent on the position of its centre of 
gravity. 
The period occupied by the earth’s axis in tracing a circle 
round the pole of the ecliptic would be about 25,000 years 
if the pole of the ecliptic were the centre of this circle, and 
the period would be about 31,000 years if the centre were 
