Page:The Moon (Pickering).djvu/64

42 the eruption of craterlet D, which, while absolutely quiescent during the whole of the earlier period, with the possible exception of one day, as shown by the figures, was decidedly active throughout the whole of the later period. Another interesting feature was the drifting of the vapour column so as to sometimes conceal and sometimes expose craterlet C. The formation of a large gap in the main vapour colvmin near the edge of the crater, in April, 1898, was something which had not previously been observed, although fine Unes crossing it had been seen occasionally dtuing the earlier period. As late as October 8, 1897, craterlet C was described, when compared with D and E, as "much the largest and most conspicuous" of the three. This had been frequently noticed and mentioned in the earller period. Upon April 6, 1898, E is referred to as "much the most conspicuous of the three," while C, although not covered by the vapour, is described as the "least conspicuous." These statements were confirmed upon April 7th and 8th. The change in the direction of the vapour cloud arising from F in the month of October, 1897, was also a marked feature of the observations. Compare figures 10 and 11.

Whether these shifting white objects are due to streams of gas issuing from the craters and carrying with them white crystals of ice, thus forming real clouds upon the Moon, or whether the crystals are deposited as soon as formed, or whether they are formed only on the lunar surface itself like hoarfrost, it is impossible to determine at present. Perhaps all three of these conditions occur. That irregularly occurring lunar changes of some sort are in progress is all that we are sure of at the present time. One can see from these few examples that the evidence in favour of the idea that volcanic activity upon the Moon is not yet entirely extinguished is pretty strong, if not fairly conclusive.

As the result of a very detailed study of the Moon in 1893, at Arequipa, a new kind of rill was discovered. From its resemblance to a terrestrial water-course it was named a riverbed, and it differs from the rills proper described in the last chapter in several important respects. In the first place, these minute rills or riverbeds are always wider at one end than at the other. Secondly, the wide end always terminates in a pear-shaped craterlet. Thirdly, their length is composed almost entirely of curves of very short radius, giving them a zigzag, winding appearance exactly resembling a terrestrial river as drawn upon a map. Fourthly, one end is nearly always perceptibly higher than the other. But here we come to a very marked distinction from the terrestrial rivers, for in the lunar rill the apparent mouth is always higher than the source. What this