220 THE BORDERLAND OF SCIENCE. 



volcanic eruptions. Our leading seismologist, Mallet, 

 lays down the rule, c without water there can be no erup- 

 tion,' and it was long since pointed "out by Humboldt 

 that all the active volcanoes of the earth are close to 

 the sea. Of course the chief evidence in favour of this 

 view consists in the nature of the substances emitted 

 during eruptions; and, in point of fact, the view may 

 be regarded as a demonstrated terrestrial relation. 

 Then it is quite impossible to conceive that so many 

 and such violent eruptions as the lunar volcanoes indi- 

 cate can have taken place without the emission of 

 quantities of vapour so enormous that a discernible 

 atmosphere would, from that cause alone, have been 

 formed around our moon. The carbonic acid gas, for 

 example, which would be poured out if the lunar vol- 

 canoes in any degree resembled ours, would form a 

 gaseous envelope of no inconsiderable depth. This 

 will be manifest when we recall Galileo's description of 

 the lunar craters as resembling the eyes in a peacock's 

 tail for number. Besides, it is difficult to imagine 

 how any planet could be formed without an atmo- 

 sphere ; and although, no doubt, the moon's small mass 

 would indicate a very inconsiderable aerial envelope, 

 yet it would not explain the complete absence of all 

 traces of air. 



The considerations here mentioned have long formed 

 one of the standing mysteries of astronomy. We see in 

 our moon a planet which ought to have oceans and 

 an atmosphere, which even would seem once to have had 

 them ; and yet she now shows no trace of either. 



