772 



The National Geographic Magazine 



Figure i 



plicity of the illustration on a globe, has 

 gained precedence over others even in a 

 land where but a small percentage of the 

 people at school ever have a chance to 

 experience the real demonstration. To 

 those who have the opportunity to ob- 

 serve for themselves this phenomenon, 

 the unaided eye does not on most occa- 

 sions 3deld very satisfactory' results. 



The experiment which forms the theme 

 of this paper is of a nature similar to the 

 ship-at-sea proof of rotundity of the 

 earth and differs most in the limits of the 

 range. The writer makes no claim to 

 originality as far as the fundamental 

 proposition is concerned. It has been 

 mentioned, more from a theoretical stand- 

 point, however, in books on astronomy 

 and geography for years. He desires to 

 show how simply the experiment may be 

 performed in regions where the proper 

 conditions prevail. It may be done wher- 

 ever a level of a mile or more is to be 

 obtained. The first choice in locality 

 falls on an inland lake, apart from tides 

 and swells, where on a calm day a level 

 and unchanging surface is available. It 

 is necessary to have three stations, one 

 intermediate in position, so that a salient 

 or an island is a requirement. This same 

 experiment may be done upon a reason- 

 ably level area by erecting rods whose 

 heights are carefully determined by a 

 surveyor's leveling instrument. 



Lake Quinsigamond, the field of the 

 experiment, is about 5J/< miles long and 

 from one-eighth to one-half mile in 

 width. It is situated along the eastern 

 boundary of Worcester, Mass. The lake, 

 like most New England lakes, is the 

 resultant of the blocking of drainage by 

 the deposit of glacial drift. Many islands 

 break the water surface, and the shore 

 line is somewhat irregular. Figure i is 

 a map of a portion of the southern end 

 of the lake. At A, a board two feet 

 sc|uare, mounted on an ordinary chart 

 standard, was located. The board was 

 divided vertically into two parts. The 

 upper part was covered with black paper. 

 This yielded a distinct horizontal line at 

 the boundary of the paper and the board. 



