exactly the reverse.' Such observations are somewhat disconcerting to the old belief, which, nevertheless, continues to maintain itself, though in somewhat modified form. It is indeed difficult, if not impossible, to explain the observed facts with regard, for instance, to the white polar caps, on any other supposition than that of the existence of at least a considerable amount of water upon the planet. These caps are observed to be large after the Martian winter has passed over each particular hemisphere. As the season progresses, the polar cap diminishes, and has even been seen to melt away altogether. In one of the fine drawings by the Rev. T. E. R. Phillips, which illustrate this chapter (Plate XX.), the north polar snow will be seen accompanied by a dark circular line, concerning which the author of the sketch says: 'The _melting_ cap is always girdled by a narrow and intensely dark line. This is not seen when the cap is forming.' It is hard to believe that this is anything else than the result of the melting of polar snows, and where there is melting snow there must be water. Such results as those obtained by Professor Pickering by photography point in the same direction. In one of his photographs the polar cap was shown much shrunken; in another, taken a few days later, it had very considerably increased in dimensions--as one would naturally conclude, from a fall of snow in the interval. The quantity of water may not be anything like so great as was at one time imagined; still, to give any evidence of its presence at all at a distance of 40,000,000 miles it must be very considerable, and must play an important part in the economy of the planet. [Illustration: PLATE XX. Mars: Drawing 1, January 30, 1899--12 hours. [lambda] = 301°, [phi] = +10°. Drawing 2, April 22, 1903--10 hours. [lambda] = 200°, [phi] = +24°. Rev. T. E. R. Phillips.] In 1877 Schiaparelli of Milan announced that he had discovered that the surface of Mars was covered with a network of lines running with perfect straightness often for hundreds of miles across the surface, and invariably connecting two of the dark areas. To these markings he gave the name of 'canali,' a word which has been responsible for a good deal of misunderstanding. Translated into our language by 'canals,' it suggested the work of intelligent beings, and imagination was allowed to run riot over the idea of a globe peopled by Martians of superhuman intelligence and vast engineering skill. The title 'canals' is still retained; but it should be noted that the term is not meant to imply artificial construction any more than the term 'rill' on the moon implies the presence of water.