250 



KNOWLEDGE 



[No^■EMBER, 1903. 



might Lave occasioued. So far, therefore, as these 

 experiments of ours go, we are not able to pronounce upon 

 Mr. Lane's, either for or against. 



Whence then did the " canals" come which were drawn 

 by the bovs of the Hospital School V 



One cause was the prolongation of dark indentations 

 invading the brighter regions. The Gehon and Hiddekel, 

 for example, on the boys' drawings were clearly jiartly due 

 to the two arms of Dawes' Forked Bay. This was my first 

 suggestion as to the prob.».ble true nature of Schiaparelli's 

 discovery, a suggestion made as long ago as 1882. But so 

 far as our experiments go 1 do not think that many of the 

 canals can Ix; explained in this way. The Forked Bay au'l 

 the two lakes Ismenius Lacus and Siloe Fous would be 

 enough to some of our boys to show them Hiddekel and 

 Gehon, but not to the majority. 



A more fruitful source of the " canals " was the intro- 

 duction of regions slightly darker or slightly brighter than 

 their surroundings. Meroe Island figured as an exam]jle 

 in the first category, Elysium as one in the second, in 

 two different experiments. And no one could wish for 

 straight er and sharper "canals'" than were drawn by a 

 good proportion of the boys to express these regions. A 

 few put in shading, fewer still put in shading and drew 

 also a dark canal as its boundary. There can, I think, 

 be no reasonable doubt that Mr. Green, when in 1879 he 

 suggested that the borders of faint shadings might have 

 given rise to some of the canals, pointed out a cause not 

 only equal to producing such an effect, Init one that does 

 so in reality. 



Another cause which proved decidedly effective was the 

 tendency of the eye to join together two small spots, 

 where these were not too small to be separately seen, by 

 a wholly subjective line. But the cause which was 

 the most effective within the limits of our experi- 

 ments with the Hospital School boys was the way in which 

 the eye summed up together minute irregular markings, 

 each too small to be separately perceived as straight 

 streaks. An examination of the drawings with reference 

 to the placing of the boys rendered this very evident. 

 At fifteen or seventeen feet the boys were near enough to 

 detect some, if not all, of the minuter details as separate 

 entities, and hence drew few canals. At thirty-five or 

 thirty-eight feet these details were for the most part too 

 small to produce any effect, even in the gross, and there- 

 fore the boys here also drew few if any canals. But 

 between these two extremes nearly every boy drew canals. 

 Halfway between them every boy without exception saw 

 " canals." For here, whilst the details were each one 

 separately invisible, they were yet capable of creating in 

 the sum a distinct impression. 



And, seen at this distance, it was very striking to see 

 how the eye, as it were "took the average" of all 

 irregularities. Tsvo irregular wavy lines drawn so as to 

 cross each other, took form as a beautiful " oasis " with 

 four straight canals radiating from it. Lines as 

 meandering as the Thames or Trent with their tributaries 

 on a map of England, straightened out as rigidly as the 

 Phison or Euphrates on Lowell's chart. If Mr. Lane's 

 effects were never secured with an entirely blank 

 '• continent," it was astonishing to see how eifectively 

 three or four dots, absolutely invisible at the distances 

 occupied, would suffice to make those effects plain. Nor 

 was it at all necessary that the dots should be put in a 

 straight line where the canal ought to run. They might 

 wander fi'om it a good deal on either side, if only tlie mean 

 line between them ran in the right direction. 



It may be objected that very little has been gained if 

 we recognise that the canals are either the edges of half- 

 tone districts, or the summation of very minute details. 



The general distribution of the true markings on the 

 ]ilanet must approximate to that shown on the charts of 

 Schiaparelli and Lowell, and the details if not straight 



Fig. 1. — Original, 

 onlv. 



Fig. 2.— C'opv of Fig. 1 at distance 

 ■ 23 feet. 



Fig. .3. —Original. " Canals " FiG. 4. -Copy of Fig. 3 at distance 

 only. 22 feet. 



Fig. 5. — Original. Irrpgnlar Fig. 6 — Copv of Fig. .'5 at distance 



lines. '33 feet. 



Fig. 7.— Copy of Fig. 5 at d 

 tance 24 feet. 



Fig. 8. — Copy of Fig. 1 at dis- 

 tance 28 feet. 



Tlie Region of the Mare Sireniim (with variations). 



lines in their ultimate conceivalile resolution are at least 

 straight lines to the eye. 



But the gain is really great. For so long as we conceive 

 of that elaborate reticulation as being a true feature of the 

 actual surface of the planet, we can hardly escape from 

 Mr. Loweirs induction. Lines so straight, so formal, so 



