26 THE BORDERLAND OF SCIENCE. 



case of one object, is much strengthened when two 

 objects of the same general aspect, but quite indepen- 

 dent of each other, both appear to be round. We 

 cannot reasonably doubt, then, that the region of space 

 occupied by each Magellanic Cloud approaches to the 

 globular form. 



But if the Magellanic Clouds are globular objects, 

 we can tell the relative limits of distance between 

 which all objects in either cloud must lie. To illustrate 

 our meaning, let us take the sun's globe. One point 

 of that globe is nearer to us than any other, and one 

 point is farther away than any other. The nearest 

 point is that which appears to lie at the centre of the 

 solar disc, the farthest would appear to occupy exactly 

 the same position, if the sun were a transparent globe. 

 Now we can tell how much farther relatively the latter 

 point is than the former, without at all considering the 

 actual distance of the sun. The sun might be only a 

 thousand miles away, or a thousand billions of miles, 

 and yet the relative distances of these two points 

 would be the same. As a matter of fact, if 'the 

 distance of the nearest point of the sun's globe is 

 called one hundred, then the distance of the farthest is 

 slightly less than a hundred and one. Precisely the 

 same reasoning applies to each of the Magellanic 



ally deduced conclusion is a just one. For instance, suppose to use 

 our ordinary modes of expression that the odds are three to one 

 against an egg-shaped fruit appearing round (under such circumstances 

 as are dealt with above), then the odds against two such egg-shaped 

 fruits appearing round would be no less than fifteen to one. 



