Geographical Maps and Sailing Charts. 



351 



miles. By reference to the map, figure 21, it will be seen that 

 the lines of measurement between the places indicated fall, for 

 some distance at least, within the inner circle, this being the 

 region of least distortion, for which the scale of miles is some- 

 what too long. By applying a correction, which consists sim- 

 ply of adding one mile to every hundred falling within the 

 inner circle, and deducting one mile per hundred for distances 

 extending beyond circle II, the corrected measurements, as 

 shown by the table, are all remarkably close to the truth, 

 exceeding two miles in but two cases. 



Group III. 





Calcu- 

 lated. 



Meas- 

 ured. 



Error. 



Cor- 

 rected. 



Error. 



Poly- 

 conic. 



Error. 



Kichniond to St. Augustine 



San Diego to Helena 



578 



997 



2529 



91 



98 



393 



961 



534 



578 



797 



1037 



1672 



583 



1004 



2535 



91 



99 



396 



968 



543 



583 



812 



1054 



1703 



+ 5 

 + 7 

 + 6 

 

 + 1 

 + 3 

 + 7 

 + 9 

 + 5 

 + 15 

 + 17 



580 



1001 



2529 



90 



98 



392 



964 



535 



580 



798 



1043 



+ 2 



+ 4 







-1 



596 

 1030 

 2565 



100 



+ 18 

 + 33 



Portland, Me. to Portland, Ore. 

 Boston to Hartford . . . 



+ 36 



+ 9 



New Haven to Albany 



Washington to Boston . . . 



o ! in 



-1 i 411 

 + 3 975 

 + 1 556 

 + 2 599 

 + 1 838 

 + 6 1086 



+ 13 



+ 18 



Key West to Pt. Isabel _ . 



+ 14 



Portland, Ore. to San Francisco 



Richmond to St. Augustine 



Vancouver to San Francisco 



Olvmpia to San Diego 



+ 22 

 + 21 

 + 41 

 + 49 



St, John, N. B. to Key West__. 



+ 31 



1681 



+ 9 



1754 



+ 82 



In this third group the measurements, with one exception, 

 are all too high, but, excepting the last three measurements, to 

 which special attention will be called, no variation amounts to 

 as much as ten miles. By referring to the map, figure 21, it 

 may be seen that the measurements are of distances extending 

 beyond circle II, this being a region on the map where distor- 

 tion is greatest and for which the scale of miles is too short. 

 By applying the correction mentioned in the previous para- 

 graph, subtracting one mile per hundred for distances falling 

 between circles II and III, and, further, two miles per hundred 

 for distances beyond circle III, the corrected values show but 

 little variation from the calculated. One measurement, Rich- 

 mond to St. Augustine, falls just within circle II, and belongs 

 strictly in Group I, with an error of five miles, but it has been 

 placed in Group III, allowing a correction of one-half mile per 

 hundred. The last three measurements are of long stretches 

 near the edges of the map, extending somewhat beyond circle 

 III, but the correction applied to them gives satisfactory 

 results. St. John, N. B., lies so far beyond circle III that the 

 correction assigned to it is somewhat uncertain. 



Thirty-three measurements are recorded in the foregoing 

 tables, and the results indicate that with a map of the United 

 States, accurately plotted in stereographic projection, very 



