of the Cuzco Valley, Peru. 





Probable 





Longitude 







error of 



Probable 



reduced by 



Weight (inversely 



Longitude 



comparison 



error in 



triangulation 



proportional to 



Station 



of field 



local time. 



to Cuzco 



probable error in 



Number. 



chronometer. 

 Seconds. 



Seconds. 



depot. 



local time). 



57 



±0.02 



±0.11 



71° 58' 38" .9 



1. 



46 



±0.04 



±0.16 



42 -8 



0.7 



363 



±0.02 



±0.11 



41 -6 



1. 



636 



±0.03 



±0.19 



54 -7 



0.6 



Weighted mean 



71° 58' 43" .4 



It is probable that these results are affected by local devia- 

 tions of the plumb line, but not so much as the latitude, for the 

 trend of the mountain ranges is east and west. 



Azimuths determined by solar compass in 1912 were checked 

 by more precise methods in 1914. 



The altitudes of the topographic map are based on the 

 height of the rails at Cuzco station, as determined by Mr. T. 

 A. Corry. As a basis for plane-table triangulation a base line 

 2 miles in length was measured along the railroad, and angles 

 from its ends were measured to points on the rim of the valley. 

 Two of these points, 6.4 miles apart, were used as bases for 

 the whole map. 



Owing to conditions incident to exploration, the quality of 

 the map is not uniform. The area marked C on the accom- 

 panying sketch (fig. 1) was surveyed by Mr. Robert Stephen- 

 son. This portion of the Cuzco Valley is intricately dissected, 

 and the time at the disposal of the topographer was brief. 

 The features are therefore somewhat generalized. The area 

 marked A is covered by a plane-table survey made by Mr. 

 Bumstead, who describes his method as follows : 



"I went over the ground very thoroughly, making many sta- 

 tions by the three point method and locating almost all the fea- 

 tures by intersection. The fact that the ground is entirely 

 unobscured by trees made this the ideal method, and levels were 

 always obtainable by vertical angles to some point or points whose 

 elevation had been previously determined. 



In a country of so great differences of elevation, the use of 

 steep vertical angles for topography is unavoidable, and unless 

 distances are known more accurately than they can be scaled 

 on a map of so small a scale, resulting determinations of eleva- 



