academy of sciences] SCIENTIFIC ACTIVITIES 185 



speculative science of geology the method of inventing hypotheses and the manner of devising 

 and applying tests for them, the implication thus given as to the previous education of our 

 geologists is not flattering. It may be added that the uncertainty in which the problem of 

 Coon Butte was left is characteristic of, not to say inherent in, the discussion of a single pheno- 

 menon, much of which lies beneath the earth's sufrace and is therefore beyond the reach of 

 observation. The probability that the cavity was produced by a subterranean expolosion — 

 such being the explanation which Gilbert appeared to prefer, as fas as he showed any preference 

 — is increased when it is noted that a number of other cavities resembling the hollow of Coon 

 Butte more or less closely are known to exist farther south in the United States and in Mexico. 2 

 The probability that the cavity was produced by the impact of a meteor — such being the explana- 

 tion which many other observers of the cavity prefer— is increased by the presence of meteoric 

 fragments near by. 



THE MOON'S FACE 



Reference has already been made to Gilbert's telescopic study of the moon at the Naval 

 Observatory in Washington during the summer of 1892, when he found Congressmen to inter- 

 fere about as much with his purpose as clouds. He wrote to a friend at that time: 



I am a little daft on the subject of the moon, being troubled by a new idea as to its craters, and I have 

 haunted the observatory for three evenings in which I have netted but one hour of observation. Clouds and 

 congressmen are about equally obstructive. Tonight I go with a stenographer so as to record what I see while 

 it is before my eye. 



Possibly it was the expense of this stenographer that was represented by an entry which 

 recurred a number of times in the pocket diary of this period: "Astronomy . . . $2.00." The 

 "new idea" — not altogether new — was that lunar craters are due to the impact of external 

 bodies and not to volcanic eruptions; the idea was very probably suggested to him by the 

 study of Coon Butte. Gilbert gave much time and more thought to the study of this highly 

 speculative problem. His "observations were practically limited to two lunations in August, 

 September, and October, 1892, a period affording 18 nights available for work"; and a period 

 in which work of this wholly nonpolitical sort may have afforded welcome distraction from the 

 disaster in which politics had involved the survey. The instrument used was a 2634-inch 

 refractor, and the magnifying power found most serviceable was 400. So persistent were his 

 visits to the observatory that he eventually traced the "terminator," or line of sunrise and 

 sunset, all across the face of the moon, and thus saw every part of its earth-turned half very 

 obliquely illuminated, so as to bring out its relief to best advantage. 



In the following winter, when Gilbert was absent on leave from the survey from January 

 to April to deliver a course of lectures at the Columbia School of Mines in New York City, his 

 inferences based on telescopic observation were tested by the experimental production of crater- 

 like forms in materials of various kinds, a laboratory at Columbia being made available to him 

 for this purpose. He dropped marbles into porridge and lumps of mud into a bed of mud; 

 he rigged up a strong sling-shot apparatus with heavy elastic bands, and thus threw pellets 

 of clay with about one-sixth of a pistol bullet's velocity into a bed of clay, varying the size and 

 velocity of the pellets and the viscosity of the clay bed; he used a large-bore musket to fire 

 balls of various materials into beds of the same materials: "A leaden ball fired into a flat sheet 

 of lead produced a cup-like crater, lined throughout with the material of the ball." The re- 

 sults of this experimental work, which the experimenter referred to in his letters as his " knit- 

 ting," supported his idea that the lunar craters are due to impact; but before the experiments 

 were»made, he had worked out what he called the "moonlet theory" to account for the craters 

 and presented it as a most alluring address when retiring from the presidency of the Philosophi- 

 cal Society of Washington in December, 1892. 3 



It has been remarked that the majority of astronomers explain the craters of the moon by 

 volcanic eruption — that is, by an essentially geological process — while a considerable number 



' N. H. Darton. Explosion craters. Sci. Monthly, iii, 1916, 417-430. [It is reported that a boring in 1925 has discovered a body of meteoric 

 iron at a depth of over 1,000 feet below the rim of Coon Butte.] 



» The moon's face, A study of the origin of its features. Bull. Phil. Soc, Washington, xii, 1893, 241-292. 



