January 5, 1905 J 



NA TURE 



227 



is made to such men as Madler, Schmidt, Webb, 

 Elger, and Nieson. 



As the result of his lunar observations in Peru, 

 Jamaica, and California, the writer has come to the 

 conclusion that physical changes do occur upon the 

 moon, and that they may be classified under three 

 heads, those due to volcanic action, those due to the 

 formation and melting of hoar frost, and those due to 

 vegetation. 



In the first class the .classical example is that of 

 Linne, which, according to the measurements of 

 Lolirmann, Madler, and Schmidt, prior to 1843, had 

 a diameter of between four and seven miles. Its 

 diameter at present is three-quarters of a mile. A 

 few years ago a new crater was announced by Klein 

 in the vicinity of Hyginus. The writer is not 

 sufficiently familiar with this region to speak from 

 personal experience, having but a few sketches of it, 

 but he believes that a change there of some sort is 

 generally admitted by selenographers. 



Perhaps no area of its size upon the moon has been 

 so thoroughly examined as the floor of Plato". It has 

 been studied at intervals of about eleven years, first 

 in 1S70 by a committee of the British Association, next 

 by A. S. Williams and others in 1S81, and again a 

 few years later, then by the writer at Arequipa in 

 1S92, and again this past summer in California. 

 In each survey about forty craterlets have been 

 mapped, and each time some new ones have been dis- 

 covered, while at the same time a few of those pre- 

 viously observed had ceased to be visible. The 

 original trigonometrical survey of 1870 was based 

 upon four craterlets located near the centre of the 

 floor, and selected as primary stations. The eastern- 

 most of these was last seen as a crater in 1888. A 

 trace of it was suspected in i8q2, but a search for it 

 this past summer with a 16-inch telescope working 

 under most favourable climatic conditions failed to 

 reveal any trace of it whatever. Even the large white 

 area upon the floor which formerly marked its posi- 

 tion has partially disappeared. 



A map of the floor of Plato, based on a survey made 

 in 1892, is given in the Harvard Annals (xxxii., 

 plate X.). On this map the craterlet numbered 3 corre- 

 sponds to craterlet number 22 of the older surveys. 

 This craterlet was tenth in order of conspicuousness 

 in 1S70. In 1881 it had risen to the seventh place. 

 In i8c)2, although carefully looked for, it could not be 

 found, and it was entered on the map as a missing 

 crater. A study of this region during the past summer 

 revealed the presence of what appeared to be a large 

 crescent-shaped bank of sand, six miles in length by 

 from one to two miles in breadth. Its height was 

 computed at not far from 1000 feet. It is the only 

 object of the sort upon the floor, and the writer has 

 so far found no previous record of its existence. 

 W"hen the sun is setting upon Plato it is by far the 

 most conspicuous object within the crater walls, and 

 was readily revealed by a 6-inch objective in Cam- 

 bridge, Mass.. working under verv unfavourable atmo- 

 spheric conditions. .At sunrise it was also in part seen 

 without difficulty under fair conditions. It seems in- 

 credible that so conspicuous an object as this should 

 have been overlooked by all the earlier observers, had 

 it then been visible. 



I accordingly wrote to Mr. Williams, and he kindly 

 sent me a list of forty-two observations made during 

 the years 1879 to i8qo, dealing with the particular 

 portion of the crater floor where this formation was 

 ■situated. Five of these observations were made during 

 that portion of the lunar day when the object is now- 

 conspicuous, and when it is much more so than any 

 of the craterlets upon the floor. Three of Mr. 

 Williams's observations record that nothing was 

 visible upon this portion of the floor. One observ- 



NO. T836, VOL. 71] 



ation records two small white spots, one of which he 

 thinks may have been the original crater, and the 

 other is possibly a neighbouring hill. Both of them 

 as shown by this sketch were evidently very small 

 objects as compared to the present formation. The 

 fifth observation records a bright streak passing 

 through the spot in question and extending for about 

 thirty miles across the floor. Evidently if the present 

 sandbank had been in existence at that time Mr. 

 Williams could not have failed to have seen it and 

 recorded it upon his sketches. Between this sand heap 

 and the crater wall a large craterlet now exists. It 

 is, in fact, the largest upon the floor, measuring about 

 two miles in diameter, but owing to its peculiar posi- 

 tion, and also to the fact that it is never bright like 

 most of the others, it can only be seen at lunar sun- 

 set, and even then is not conspicuous. 



Turning now to the second class of physical changes 

 visible upon the moon, those due to the formation and 

 disappearance of hoar frost, we find numberless ex- 

 amples scattered over the surface, but in most cases 

 favourable atmospheric conditions and a large glass 

 are necessary to render them clearly visible. Before 

 dealing with any specific cases, however, it may be 

 well to endeavour to answer some of the objections 

 raised on theoretical grounds to the possibility of the 

 existence of water vapour upon the moon. 



The writer believes that he himself was one of the 

 first to point out that if water vapour existed upon the 

 lunar surface, it must sooner or later be dissipated 

 into outer space {Astronomy and Astrophysics, 1892, 

 xi., p. 781). That such a dissipation must have been 

 going on in times past seems to be inevitable, but 

 before reaching a conclusion as to the present existence 

 of water vapour upon the moon, there are one or two 

 important considerations that must be taken into 

 account. 



Vulcanologists are now generally agreed that the 

 vast quantity of water, amounting to thousands, and 

 sometimes to millions of tons, given off during 

 volcanic eruptions is not rain water, nor yet water 

 that has reached the interior from the ocean, but is 

 water that either is being expelled for the first time 

 from the earth's interior or is being expelled by 

 heat from the rocky materials of the earth's crust with 

 which it was previously united by the forces of crystal- 

 lisation. If the earth is still discharging such large 

 quantities of water from its interior there is no reason 

 why the moon should not be doing the same thing. 

 It is true the moon is smaller, but then also it began 

 life later than the earth. The reason why the earth 

 has oceans is that it is large enough and massive 

 enough to retain the expelled water in that form. The 

 mooii, on the other hand, is too small to do so, and 

 the water therefore appears scattered widely over its 

 surface in the form of hoar frost before being dissi- 

 pated into outer space. 



Another objection to the theory of the existence of 

 water vapour that has been raised is the statement 

 that there is no evidence of erosion upon the moon. 

 This statement is clearly a mistake, but the eroded 

 valleys are small, and it requires good atmospheric 

 conditions to detect them. Fairly conspicuous ex- 

 amples, however, exist upon the central peaks of 

 Theophilus and Eratosthenes. Although the valleys 

 are small, it is hard to understand how the compara- 

 tively minute amount of hoar frost at present found 

 in these regions could have produced so great an 

 effect, and we must conclude that formerly there must 

 have been a great deal more of it. The only strong 

 evidence that water in the liquid state ever existed upon 

 the surface of the moon lies in the dry river-beds. The 

 best example of these lies on the eastern slopes of 

 Mt. Hadley, at the base of the Apennines. Another 

 river-bed, partially fragmentary, discovered this past 



