And now for the views of Professor Lowell himself with regard to the doubling of the canals. From his observations, he considers that no pairs of railway lines could apparently be laid down with greater parallelism. He draws attention to the fact that the doubling does not take place by any means in every canal; indeed, out of 400 canals seen at Flagstaff, only fifty-one--or, roughly, one-eighth--have at any time been seen double. He lays great stress upon this, which he considers points strongly against the duplication being an optical phenomenon. He finds that the distance separating pairs of canals is much less in some doubles than in others, and varies on the whole from 75 to 200 miles. According to him, the double canals appear to be confined to within 40 degrees of the equator: or, to quote his own words, they are "an equatorial feature of the planet, confined to the tropic and temperate belts." Finally, he points out that they seem to _avoid_ the blue-green areas. But, strangely enough, Professor Lowell does not so far attempt to fit in the doubling with his body of theory. He makes the obvious remark that they may be "channels and return channels," and with that he leaves us. The conclusions of Professor Lowell have recently been subjected to strenuous criticism by Professor W.H. Pickering and Dr. Alfred Russel Wallace. It was Professor Pickering who discovered the "oases," and who originated the idea that we did not see the so-called "canals" themselves, but only the growth of vegetation along their borders. He holds that the oases are craterlets, and that the canals are cracks which radiate from them, as do the rifts and streaks from craters upon the moon. He goes on to suggest that vapours of water, or of carbonic acid gas, escaping from the interior, find their way out through these cracks, and promote the growth of a low form of vegetation on either side of them. In support of this view he draws attention to the existence of long "steam-cracks," bordered by vegetation, in the deserts of the highly volcanic island of Hawaii. We have already seen, in an earlier chapter, how he has applied this idea to the explanation of certain changes which are suspected to be taking place upon the moon. In dealing with the Lowell canal system, Professor Pickering points out that under such a slight atmospheric pressure as exists on Mars, the evaporation of the polar caps--supposing them to be formed of snow--would take place with such extraordinary rapidity that the resulting water could never be made to travel along open channels, but that a system of gigantic tubes or water-mains would have to be employed!