310 FROM NEBULA TO NEBULA 



4. The fact that the canals appear on the dark blue 

 regions as well as on the ochre leads him to assume that 

 both are land ; the former, in his opinion, being probably 

 the basins of evaporated oceans and now covered with 

 vegetation. 



5. The canals follow the arcs of the planet's great 

 circles, so that we who look centrally down upon them see 

 apparently straight lines. This studious regard for 

 choosing the shortest distance, he opines, implies not only 

 provident economy in design, but also a very high degree 

 of technical knowledge and skill. Necessarily, there are 

 many points of intersection, and, strange to say, at the 

 majority of these there are circular dots about 75 miles 

 in diameter, which in color correspond with the canals. 

 These Mr. Lowell conceives to be oases, Martian cities, as 

 it were, environed by irrigated farms. Not this alone, 

 these canals connect with certain caret-shaped spots that 

 appear to be the "salient points" of the blue-green 

 patches ; and from certain of the oases canals branch out 

 numerously, always in straight lines, to other oases, 

 forming a sort of open network. They always lead to 

 definite destinations, never stop short as rivers might do, 

 and, again unlike rivers, they preserve a surprising 

 uniformity of line throughout. 



6. In the first days of summer of each hemisphere 

 these canals, he states, begin to grow in distinctness, 

 starting at the polar cap soon after it commences to thaw 

 and thence gradually continuing down to the equator. 

 At such times, too, certain of the canals have a trick of 

 pairing or doubling, technically known as "gemination". 



7. The aggregate length of the canals is enormous, 

 probably as much as 40,000 miles, or three times the cir- 

 cumference of the Martian globe. Five hundred and 

 twenty-two of them have been mapped, the shortest being 

 not less than 250 miles in length and the longest, the 

 Eumenides-Orcus, stretching the enormous distance of 

 3450 miles. At least a dozen extend to points far withih 

 the polar circles, and would no doubt be found to reach 

 even to the pole itself, were their ends not obscured by 



