THE DOCTRINE OF A ROUND EARTH 67 



the time of an eclipse required that the sun be larger than the 

 earth ; for if the earth is a globe the shadow it throws will be, 

 if not a cylinder, a cone, either upright or inverted, as the 

 source of illumination is larger or smaller than the earth. Here 

 is doubtless another diagram that Bion and others who thought 

 about it must have drawn : 



FIG. 2. 



If the sun were small and near, as at S, then the shape of 

 the earth's shadow would be that indicated by the lines S A, S A', 

 and the moon in travelling round the earth would be blotted 

 out for many hours during an eclipse. This is not the case. 

 Often only a part of the moon's disk is cut off during an eclipse ; 

 never does it remain in the earth's shadow much more over three 

 hours. Occultations of the sun make it clear that it is vastly 

 larger than the moon : it must also be vastly larger than the 

 earth in order to realise the conditions indicated in the figure. 



One step further. If the sun be larger than the earth, then 

 its distance must be immeasurably vast, for when its rays strike 

 the earth they are to all intents parallel. That is attested 

 by the fact that the shadows cast by sunlight alike of the 

 Pyramids and of a knife-blade, are sharp. And again, if the 

 sun were not at a great distance, at points on the earth near 

 the poles (P P in the figure), a part of the sun's disk would be 

 cut out by the curve of the earth continuously. 



On the other hand, if the sun is larger, and not at a very 

 great distance, more than half the earth would be continuously 

 illuminated, and we should never have equinoxes and equal 

 days, as we do, nor, for example, would the poles have six 

 months day and six months night as Bion taught. 



Thus, as it chances often, the elucidation, the bringing into 



