1862.] 



The Trigonometrical Savvey of India. 



37 



67. Process of estimating terrestrial refraction. — Let A, B, C, 

 D, (vide figure) be plain stations, T and S stations on the Sub- 

 Himalayas, and I. to IV. inaccessible points on the range of perpetual 

 snows. Let the values of f at T and S equal respectively f and^. 

 We may deduce from these, two trustworthy values of the heights 

 of I. and II. Calling this mean height of I = I m , and re- 

 membering that we have elevation (E) at C of I, as also the 

 contained arc for C I == (c) given, it is clear that the values of f at 

 C, corresponding to I m may be found. Let this value =f Pro- 



f z -{_ f 2 _i_ . . . -h f 



ceeding in the same manner we shall find f c = ' — — 



° n 



Similarly fa &c, may be obtained, and withj^, fo &c, may be com- 

 puted III,„, IV m &c, from which again in turn may be found the 

 values of f for the other plain stations from which III, IV &c , have 

 been observed. By this process the computed values of f are deter- 

 mined nearly in terms of ft andy^, errors of observation not being 

 taken into account. It remains to mention how f s and f t were 



obtained. 



68. The computations originate from Senchal and Tonglo hill 



stations, at which stations, the. following mean value of f was in the 



