330 FROM NEBULA TO NEBULA 



to discover any rain clouds in the lunar atmosphere. 

 Indeed, even in this matter of atmosphere the moon ap- 

 pears to be no less destitute, the proof whereof consists 

 in the circumstance that when stars are occulted by her, 

 as occasionally happens, they disappear and reappear 

 suddenly without exhibiting any signs of atmospheric 

 refraction. 



This poverty of the moon in water and air may be 

 due to one of four reasons; first, to a scantiness in the 

 original endowment of the body in these two elements, 

 or, second, to the subsequent escape of their swifter mole- 

 cules, as suggested by Doctor Stoney, or, third, to per- 

 colation and absorption into the moon's interior, or, 

 fourth, to transformation into another state or substance 

 wliereby they have become masked and unrecognizable. 

 Of these four hypotheses, it seems, contemporary scien- 

 tists have finally settled upon the second as the nearest 

 approximation to the truth yet proposed (though still 

 admittedly far from satisfactory) while the last, and, as 

 I hold, the true one, seems never heretofore even to have 

 been thought of, much less discussed. 



The most remarkable thing about the strange mark- 

 ings on the lunar ocean beds is not the fact of their exist- 

 ence, nor even the peculiarity of their shapes, but the 

 singular resemblance which they bear to the extraordi- 

 nary structures that exist in such profusion on the con- 

 tinental areas. The same forces that modeled the first 

 class must have unquestionably modeled the second t 

 The only substantial difference between the two orders 

 is one of scale, those on the continents vastly exceeding 

 the others in height; although there exists this minor 

 distinction, that the configurations occupying the bottoms 

 exhibit more of a ruinous appearance. 



Compared to the earth's, the lunar surface is ever 

 so much rougher, and these roughnesses differ so funda- 

 mentally in their general character as to compel the con- 

 viction that they owe themselves to radically different 

 modeling forces. Let us, if you please, examine some of 

 the principal of these 



