522 



SCIENTIFIC NEWS. 



[Nov. 23, iS 



up to a certain latitude by the annual summer presenta- 

 tion of either hemisphere to the sun." ("Fuel of the 

 Sun," page 169.) 



Now let the reader turn to page 433 of this magazine, 

 and he will see by Schiaparelli's diagram that what I 

 described theoretically twenty years ago was observed 

 by him in May and June of this year. The letters bbbbb 

 surrounding each picture indicate " white spaces," besides 

 the well-known polar patches. 



With apology for inevitable egotism, I will quote my 

 own essay again. " At the poles and for some distance 

 around them, the annual amount of deposition must 

 exceed the annual amount of thawing and evaporation, 

 and therefore a gigantic glacial mountain must there 

 accumulate, with a continual growth and tendency to 

 assume the conical form. As the deposition of ice 

 crystals would commence before actual sunset, and 

 would probably reach its maximum, or even be finished, 

 before reaching the boundary line of day and night (in 

 consequence of the thinness of the atmosphere and the 

 resulting rapidity of radiation), the building up of this 

 polar mountain would be very irregular. In mid-winter 

 the lower slopes would receive the greatest accessions. 

 With the advancing line of daylight the elevation of the 

 zone of maximum deposition would increase until it 

 reached the summit. This coincidence of maximum de- 

 position with the summit would occur twice a year, 

 before and after mid-summer. During the summer the 

 only regions receiving any deposition at all would be the 

 summit and its immediate vicinity ; while at the same 

 time the sides would be rapidly thawing by the power- 

 ful action of the continuous sunshine of the one long 

 arctic summer day. At this season, the slopes of the arctic 

 mountain would be riven by gigantic ice-floods and water- 

 floods, avalanches, glaciers, and torrents." 



The " canals " are, I believe, these " water-floods " or 

 " torrents." Their theoretical magnitude will be 

 understood by considering the dimensions of the polar 

 ice mountain, which, according to Professor Phillips, 

 reaches in winter to latitude 50, that is, nearly half 

 way from the pole to the equator of the planet; the 

 polar snow circle as seen by the telescope has a diameter 

 of 2,000 miles in winter, on a world having but half the 

 diameter of ours. Its reduced summer diameter is about 

 500 miles; thus a ring of ice measuring 6,283 miles inouter 

 circumference, and 750 miles broad, is thawed away in 

 half a Martian year. Its thickness, according to my 

 original estimate, is measurable in miles. The conse- 

 quent floods must be enormous, vastly beyond anything 

 within the reach of our terrestrial experience, although 

 on a globe so much smaller, only one-fourth of our area. 



Over what kind of ground would these great torrents 

 flow ? As I have shown in the above quoted essay, the 

 seas of the tropical and temperate zones of Mars cannot 

 be of liquid water ; they must be oceans of ice, the mere 

 surfaces of which are daily thawed and refrozen. Thus 

 the apparent paradox of these rivers crossing the seas as 

 well as the land, is explained. 



Besides this, there should be, as I have argued, a great 

 moraine heap surrounding the regions of polar snows ; 

 this ring moraine " consisting of the materials which the 

 advancing polar glacier has scooped out and driven before 

 it. This perpetual out-thrust of the great polar glacier — 

 this continual erosion of the circumpolar regions of the 

 planet, ever at work since the date of its primeval con- 

 solidation, must have produced a sensible alteration of 

 the shape of the planet — must have flattened it in the 



immediate neighbourhood of its poles, and bulged it in 

 the moraine regions beyond." 



We thus obtain theoretically a great ring-shaped cir- 

 cumpolar lake, walled up by a surrounding ring-shaped 

 ridge of moraine matter of mountainous dimensions. 



Between the time of the writing and publishing of 

 these speculations, Mr. Browning published his stereo- 

 grams of Mars and some charts. On- these are shown a 

 ring of water completely surrounding the limits of the 

 North Polar ice, to which Mr. Proctor gave the name of 

 Schroter Sea, and beyond this a circular ridge which he 

 calls " Laplace Land." Similar features are repeated 

 around the South Pole, where the circular erosion valley 

 is named " Phillips Sea," and the ring moraine is, on its 

 polar boundary, so symmetrical that it nearly represents 

 a parallel of latitude, but on its equatorial side has some 

 remarkable promontories bearing the names of " Cassini 

 Land " and " Lockyer Land." 



If my theoretical views are sound, the waters of these 

 circumpolar seas must in the Martian summer time rise 

 to the upper limits of the great moraine barrier and over- 

 flow or break through it, then flow onward in a course 

 determined by the general slope, which should be pretty 

 regular, owing to the regularity of the piling-up action. 



In reference to this it should be remembered that the 

 Martian summer is nearly double the length of ours, the 

 tilting of its axis a little greater, and the eccentricity of 

 its orbit much greater, all increasing the variations of 

 climate. Therefore the winters of the polar regions must 

 be intensely cold and the summers hot, and their heat 

 most effective, on account of the continuous sunshine. 

 The amount of thawing would therefore be enormous, and 

 floods pouring over or through the barrier proportionate. 



If it should happen that two notches or outlets of the 

 great moraine ridge occur near together, a great river 

 would flow from each. With similar conditions of slope, 

 their courses should be nearly parallel; hence the "germi- 

 nation" or twinning of the "canals." I must not be tempted 

 into further details, having already overrun my space. 



CAUSE OF THE ABSENCE OF TREES 

 IN THE AMERICAN PRAIRIES. 



"pOR a considerable time the grass-fires which in 

 former days recurred almost periodically have 

 been regarded as the chief cause of the absence of trees 

 in the wide prairie districts of North America, since 

 they must infallibly destroy any young growth. Still, 

 they cannot have been the original cause, for the prairies 

 existed before human beings and fires could have 

 occurred. Another hypothesis has sprung from recent 

 investigations on loess and laterite ; the " ground-water " 

 in the prairie regions lies at such a depth as to be out of 

 the reach of the roots of the young trees. Hence seed- 

 lings perish in the first prolonged drought, and trees can 

 in consequence maintain themselves only on the margin 

 of rivers. One striking phenomenon, however, could 

 not be thus explained — the occurrence of isolated woods, 

 not in hollows, but, on the contrary, on sandy eminences 

 where they might have least been expected. The most 

 extensive case of this kind is the forest belt known as the 

 Cross Timbers. 



Professor Thomas Meehan has recently given a very 

 satisfactory explanation of this phenomenon in a paper 

 read before the Philadelphia Academy of Sciences. He 

 has observed in the neighbourhood of Roan Mountain, in 

 North Carolina, small tracts of grass land lying isolated 



