considered. Meteors falling into the Sun would make it grow bigger: we believe, on the contrary, that it is getting smaller—"shrinking with the cold." That notion was mentioned in the last lecture in connection with the planets, and we saw how it led to the further notion that they would spin more quickly and ultimately throw off satellites. If, as we suppose, the planets themselves are satellites of the Sun thrown off in the same way, then at any rate he must have been shrinking in time past, and we know of no reason why the shrinkage should have stopped. But it seems extraordinary that we should try to explain his generous output of heat by saying that he is "shrinking with the cold": it looks as though we are explaining a thing by its opposite. We can, however, see how this may happen. Supposing your father were to come home one day and say to your mother: "I have been balancing accounts, and I find we have spent a great deal of money this year; we must reduce our establishment; let us give up our motor-car"; it might then happen that giving up the motor-car saved more money than was needed to bring the expenses down to the right figure, so that in other ways the family would be richer than before. It is rather like that with the Sun: losing a lot of heat makes him reduce his establishment—makes him shrink, but the very shrinking causes a development of heat—more heat, perhaps, than he lost, so that he is actually hotter after the shrinkage than before.
This way of getting heat by shrinkage does not appeal to our imaginations so readily as the idea of
