THE MOON IN MASQUERADE 121 



show an exceedingly rough topography diametrically 

 contrasting with the natural one. The surface of the 

 snow is therefore necessarily deeply cut and furrowed. 

 When the daily thaws come on, the melting (not evap- 

 oration ) takes place, as a matter of course, only on 

 the surfaces directly exposed to the sun's rays; but the 

 water, being many times heavier than an equal bulk 

 of snow, instantly abandons the sharp pinnacles and 

 crests, leaving them pure white, but sops the furrows, 

 which anyhow are in shadow, thus creating the general 

 appearance of white ray a ayainst a drab background of 

 xlusft, and thereby incidentally explaining another of 

 the lunar mysteries that has long resisted solution. 



Naturally the melted water that manages to seep 

 back into the basins must, while the sun is still high, 

 immediately undergo again the process of evaporation. 

 Owing however to the frigidity of the planet's quiescent 

 atmosphere, as well as to its own self-built prison walls, 

 the rising mist changes into snow before it escapes from 

 the caverns, in which condition it is of course unde- 

 tec table as vapor. Some of this snow spreads promis- 

 cuously over the planet's surface, but a proper share 

 of it remains to restore and heighten those very walls. 

 This piling-up process cannot go on indefinitely with- 

 out eventually causing the peaks to telescope upon 

 themselves from time to time, partly on account of 

 their own overgorged weight and partly because of 

 the undermining of their bases by the periodical flood- 

 ing of the bottoms. This telescoping process gives us 

 the key to the terraced effect so generally observed on 

 the inner sides of the crateral ramparts, and supplies 

 the reason for the otherwise surprising precipitousness 

 of their faces. It also explains why the mountains and 

 ramparts automatically preserve a maximum uniform- 

 ity of height. 



