42 A MANUAL OF TOPOGRAPHIC METHODS. 



BASE-LINE MEASUREMENT. 



This is, ordinarily, the first of the preparatory steps toward map making. 

 Upon the proper selection of the site of the base line and its correct meas- 

 urement depends all the subsequent work of tri angulation. The site must 

 be reasonably level. It is not essential that it be absolutely so, but the 

 more closely it approaches a plane the less difficulty will be experienced in 

 making an accurate measurement, The site should afford sufficient room 

 for the measurement of a base from 5 to 10 miles in length. A base less 

 than 5 miles in length is not an economical one, inasmuch as it is less 

 costly to extend the base than to complicate the expansion. A greater 

 length than -10 miles is unnecessary, because this length permits of easy 

 expansion, and, if the length be greater than this, it may be difficult to con- 

 struct signals at the two ends of the base, which will be intervisible. 



The ends of the base must be intervisible, and they must be so situated 

 with regard to suitable points for expansion and triangulation as to form 

 well proportioned figures. Whenever possible, the base line should form a 

 side or diagonal of a closed quadrilateral or pentagonal figure. 



While it is unnecessary to devote time to obtaining the extreme of 

 accuracy in the measurement of a base, this measurement should be so 

 accurate that its errors can not affect the map, although multiplied many 

 times in the associated triangulation. All necessary precaution should be 

 taken to secure this result. 



Various methods and instruments have been employed in the measure- 

 ment of base lines upon the Geological Survey. At first wooden rods were 

 employed, varnished and tipped with metal. When used in measuring, 

 these were supported upon trestles and contacts made between them, with 

 considerable refinement. The advantage of using these rods consisted in 

 the fact that their length is but slightly affected by temperature, which is 

 the main source of error in base-line measurement, and being thoroughly 

 varnished they were not greatly affected by moisture. 



Subsequently bars of metal were employed of the pattern known as 

 the Coast Survey secondary bars. These consist each of a steel rod between 

 two zinc tubes. As the two metals expand at different rates under changes 

 of temperature, their relative lengths at any temperature. as compared to the 



