386 SURVEY OF THE COAST 



chances of forming the subsequent triangles will not onlv 

 accelerate the work by the greater number of points which 

 may be determined from them, but also enable the operator 

 to choose the most advantageous combination of triangles: 

 and if the survey shall extend over a whole country, it is 

 proper to begin in sucli part of it as will soon lead to the 

 largest triangles possible. 



This was the principle which induced me to begin in the 

 neighbourhood of New York, after 1 had received a ktter 

 from Mr. Dallas, Secretary of the Treasury, authorising me 

 to begin wherever I should think it most advantageous. 



The configuration of the country, particularly the direction 

 of some principal chains itf mountains, may guide us in the 

 research for a base line, because it is generally parallel with 

 straight ranges of mountains, that the more level and exten- 

 sive plains are found, and the first triangle point will fall 

 upon some elevation in an opposite range of mountains, if 

 these be not very elevated. 



It is therefore necessary to make first a preliminary recon- 

 noitring survey of such a tract of the country at least as may 

 be likely to furnish the base, and a complete system of tri- 

 angles, admitting of verification. The more minutely this 

 can be done, the more advantageous on every account will 

 it prove for the actual survey. 



In an uniform simple triangulation, it is most proper to 

 place the two base lines, which are necessary at least in all 

 such works, near the two extremities of the triangulation, 

 and therefore to make also at the beginning the necessary 

 reconnoitring for that purpose. But in my casein this work, 

 it was desirable to obtain in an early staseoftlie work a ve- 

 rification of the linear unit upon which the triangle was 

 founded, to allow me to make use of it for the detail sur- 

 veys as soon as possible. Besides, in all cases where the 

 triangulation is to serve for geographical purposes, it may 

 most generally be expected that several base lines will be 

 measured in the course of the work. 



This first reconnoitring requiring the most extensive and 



