324 HYDHOGRAPHir WORK OF U. S. GEOLOGICAL SURVEY 



bright sunsliine into coni[)ar:itive darkness its power of estimating 

 intensity of illumination is so <listurl)eil as to ^\\'e rise to great seem- 

 ing fluctuations — that is, a constant illuniinatinn will for a time seem 

 irregularly variable. It is possible, though not ver}' probable, that 

 the seeming bands may be due to this cause, and may therefore have 

 no physical reality. 



But a possible i)hysical cause for liglit fluctuations on a sheet is not 

 far to seek. The twinkling of a star shows that if a star shone brightly 

 enough to illunnnatc a lai'ge white sheet we should see the illumina- 

 tion t«i be a constantly chnngingand fiickering one, varying in a way 

 too irregular to admit of exact (h^scri])tion. Possibly the same m:iy 

 be true of the light cast by the very thin crescent of the sun just be- 

 fore and after the total phase of an eclipse. 



If this be the true explanation the appearance of well-defined bamls 

 of detinite breadth will still remain unexplained. But thi.s discord- 

 ance idread}' mentioned seems to show that this assignment of precise 

 forms is of the nature of an o|)tical illusion. Illusions of this sort are 

 so common and so easy to fall into, we miglit almost say unavoidable, 

 that no improl)ability attaches to them. 'I'he more careful and exact 

 an observer is, the more likely he is to detect them in his own case 

 and the less confi<lence he will have in his own ol)servations of such 

 phenomena as those in (piestion. 



HYDROGRAPHIC WORK OF THE U. S. GEOLOGICAL 



SURVEY 



In tlie .state of New York much intere.st has been aroii.seil in tlie question of 

 the water supply of tlie state, botii because of its importan<'e for supplying- 

 power to manufactories and because of the proposed deep waterway from I lie 

 Great Lakes to the ocean. Owing to the scarcity of water, some of ttie liiglier 

 levels of the Erie Canal are operated with difficulty in tlie <h'y seasons. For a 

 larger canal much more water must be had. How this extra supply may he 

 obtained requires careful examination conducted through several years. On 

 this problem the Hydrographic Oivision of the U. S. Geological Survey is now 

 at work, cooperating with state oliicials of New York. The greater part of the 

 water comes from the Adirondack Mountain area, where the forests are being 

 preserved largely for their beneficial influence in water conservation. The ev- 

 tentof this influence is also heing investigated, in order to determine, if possible, 

 the value of the forest in regidating tlie stream flow. 



The long continueii drought in California has forced i)ublic attention to the 

 impeiative necessity of providing water storage for the irrigation of farms and 



