July, 1893. THE SURFACE OF THE MOON. 49 



diameter of 800 miles. There is no important discrepancy in the 

 vertical dimensions. Lunar craters of the first rank range from 

 8,000 to 15,000 feet in depth ; terrestrial craters, probably from 2,000 

 to 4,000 feet. The contrasts as to form are of greater importance, 

 and are all fully set forth in the paper. Professor Gilbert concludes 

 by observing that " the volcanic theory, as a whole, is therefore 

 rejected, but a limited use may be found for the maar phase of vol- 

 canic action [no eruption of lava, but merely an explosion of steam] 

 in case no other theory proves broad enough for all the phenomena." 



Passing next to the tidal theory, Professor Gilbert puts aside the 

 views of Rozet, Hooke, Bergeson, and Humphreys, and considers 

 those of Faye, Ebert, and Hannay. These postulate a time when 

 the moon was liquid, with the exception of a thin crust. The moon 

 then rotated more rapidly than now, and great tides, excited by the 

 earth's attraction, racked and cracked its crust, and here and there 

 squeezed out a portion of the liquid nucleus, which flowed back again 

 when the tidal wave had passed ; but congelation caught the flood at 

 its edges, so as to mark its limit by a solid ridge. By each successive 

 tide the operation was repeated, with the result that the wall was 

 given a circular form, and was gradually built up. The process was 

 finally closed by the congelation of lava in the orifice, and while 

 congelation was in progress the last feeble eruption sometimes 

 produced a central hill. Professor Gilbert asks, in criticising this 

 theory, for an explanation of the multitude of small craters overlying 

 the larger, and why, if the crust were divided by fissures, would not 

 the tensile strains wrought by the crest of the tidal wave cause the 

 fissures to gape, instead of forcing out the liquid through apertures 

 here and there ? Or, if there were no fissures, would not the strains 

 suffice to produce them ? He further points out that there are 

 numerous craters, of small or medium size, occupying slopes of the 

 greater crater rims, and the initiation of these by tidal process seems 

 impossible. Whatever lava escaped from an orifice on a slope would, 

 he remarks, flow down the slope, instead of being drawn back. 



The views of Ericsson and Peal, that the site of each crater 

 was once occupied by a pool of water which, by heat from below, 

 was vapourised, converted into snow, and from which, was eventually 

 accumulated an annular ridge, he criticises as follows : — If the 

 rim were built up by the quiet fall of an infinitude of ice particles or 

 snow flakes, its configuration should be smooth and regular instead 

 of exhibiting the rugosity actually observed. The postulated heat of 

 the central area might render the inner slope steep, and even produce 

 the inner cliff and terraces, but the theory affords no explanation of 

 the wreath nor of the central hill ; it fails also to account for the 

 small craters formed on the rims and slopes of the larger, for the 

 bottoms of these are far above the assumed rock plain of the moon, 

 through which the theory supposes the internal heat to have been 

 communicated. Professor Gilbert admits that he is unable fully to 



D 



