PAPERS PRESENTED AT GENERAL SESSIONS 59 



comes of this energy after it has assumed the form of 

 radiation? Shall we think of it as spending an eternity 

 upon a voyage which admits of no other incidents than a 

 veiy rare reflection from or absoi^ption by some physical 

 object with a consequent reduction in wave length? Or 

 shall we think of the radiant form of energy as but a 

 phase in a grand cycle of existence? Obviously such ques- 

 tions are fundamental ones in a rational cosmology. We 

 cannot hope to answer them directly from experience. 

 The best that we can do is to frame our answers in accord- 

 ance with some chosen system of postulates and to test 

 them by their harmony or discord with our astronomical 

 and our physical knowledge. 



To most people it seems desirable to think of space as 

 euclidian and infinite ; the physical universe as infinite in 

 extent, and such portions of it as come under our observa- 

 tions not essentially peculiar as compared with other por- 

 tions; that our epoch of time has nothing essentially 

 unique about it; and that, considered in a sufficiently 

 large way, the physical universe is not tending towards 

 some state which is inherently different from that with 

 which we are familiar. Finally we recognize that the 

 physical universe is organized into physical units, such as 

 atoms and molecules, crystals and cells, stars and galaxies, 

 each of which is an organization of smaller units endowed 

 with a certain amoimt of energy and possessing its own 

 peculiar properties. We have no reason for supposing 

 that the electron is the smallest physical unit, nor, indeed, 

 that there is any smallest physical unit whatever — and 

 similiarly that there is no largest one. If we accept these 

 postulates then our cosmology, if it is rational, must be 

 consistent with them and at the same time in harmony 

 with all of the known facts in every domain of physical 

 science. As the hypotheses of physical science are rarely, 

 if ever, in complete accord with all of the known facts, 

 even when the field covered by the hypothesis is a rela- 

 tively small one, it is hardly to be expected that the hy- 

 potheses of comology will be of a different character. 



Consider for a moment a dilute solution of a chemical 

 salt in water. If it were impossible to obtain concentrat- 



