LIFE IN SPACE AND TIME 



as to make repetition superfluous. But, though the raw 

 materials for these: carbon, oxygen, hydrogen, are wide- 

 spread, it cannot be expected that the compounds them- 

 selves will be found everywhere. It is well known that a 

 celestial body of small mass, such as the Moon, cannot 

 permanently retain an atmosphere. The molecules of a 

 gaseous atmosphere are in rapid motion, and flying about 

 in all directions, and only the gravitational attraction of 

 the body keeps them from diffusing away into the practically 

 perfect vacuum of interplanetary or interstellar space. 

 For a body no larger than the Moon (and at ordinary tem- 

 peratures), gravity is not strong enough to keep them back, 

 and an atmosphere, even if one were artificially supphed, 

 would escape into space. Carbon dioxide, being gaseous, 

 would escape (though more slowly than the hghter gases) 

 and water-vapor would be readily lost, so that hquid water, 

 too, would gradually evaporate away. 



The Earth's attraction is abundantly sufficient to retain 

 an atmosphere, and as we shall see later. Mars is also able 

 to do so. For bodies roughly similar to the inner planets 

 of our system, the hmiting size, below which an atmosphere 

 and ocean cannot be retained, appears to correspond to a 

 diameter of about 3000 miles. We may expect smaller 

 bodies to be airless, waterless, and hfeless, while on those 

 considerably larger, water and carbon dioxide are hkely to 

 be ubiquitous. 



(c) A Surface Temperature Below the Boiling Point of 

 Water at All Times, and Above the Freezing Point at Least 

 at Regular Intervals. It is a commonplace that liquid water is 

 a necessity for the growth and reproduction of hfe, though 

 not for its mere existence in a dormant state. Terrestrial 

 experience shows that life, however, may persist under 

 alternations of temperature with a very low minimum, 

 provided that the maximum temperature is above freezing 

 and lasts long enough. If this condition is met, the lower 

 Hmit of the fluctuation could probably descend far below 

 any observed terrestrial temperatures without making Hfe 

 impossible. The upper hmit appears at first to be more 

 sharply defined, for even temporary boihng appears to 

 be fatal, and most terrestrial forms succumb at temperatures 



