ASTRONOMY. 251 



tern ; secondly, that its light and heat would bt; imparted to 

 the other planets much more irregularly than light and heat 

 are now received from the sun. 



{*) II. Another thing, in which a choice appears to be 

 exercised, and in which, among the possibilities out of which 

 the choice was to be made, the immber of those whicfi were 

 wrong bore an infinite proportion to the number of those 

 which were right, is in what geometricians call the axis of 

 rotation. This matter I will endeavor to explain. The 

 earth, it is well known, is not an exact globe, but an oblate 

 spheroid, something like an orange. Now the axes of rota- 

 tion, or the diameters upon which such a body may be made 

 to turn round, are as many as can be drawn through its centre 

 to opposite points upon its whole surface ; but of these axes 

 none are iiermanent, except either its shortest diameter, that 

 is, that which passes through the heart of the orange from 

 the place where the stalk is inserted into it, and which is 

 but one ; or its longest diameters, at right angles with the 

 former, which must all terminate in the single circumference 

 which goes round the thickest part of the orange. The 

 shortest diameter is that upon which in fact the earth turns, 

 and it is, as the reader sees, what it ought to be, a perma- 

 nent axis ; whereas, had blind chance, had a casual impulse, 

 had a stroke or push at random set the earth a spinning, 

 the odds were infinite but that they had sent it round upon a 

 wrong axis. And what would have been the consequence? 

 The difference between a permanent axis and another axis 

 is this : when a spheroid in a state of rotatory motion gets 

 upon a permanent axis, it keeps there ; it remains steady 

 and faithful to its position ; its poles preserve their direction 

 with respect to the plane and to the centre of its orbit ; 

 but while it turns upon an axis which is not permanent— 

 and the number of those we have seen infinitely exceeds the 

 number of the other — it is always liable to shift and vacil- 

 late from or\Q axis to another, with a corresponding change 

 in the inclination of its poles. Therefore, if a planet once 



