Review of the Results of the U. 8. Coast Survey. 77 
inclinations, and flexures of the measuring bars, and the con- 
tacts of their extremities must be observed with exquisite nicety. 
Here at the very first step in the work, the science of physies 
is called upon for its indispensable aid, and the reflecting pyro- 
meter measures the expansions of the bars with an accuracy 
which is almost without limit. Delicate levels give the corree- 
tions for the inclinations and flexures, while the contact level de- 
termines the contacts of the successive bars. To such a degree 
of perfection have the measurements been brought that the 
probable error in determining the length of a base five miles in 
length does not under favorable circumstances exceed a. few 
tenths of an inch. The base line being measured, the triangula- 
tion begins. Signals at distances of ten to twenty miles from 
either end of the base are observed in suecession, and their an- 
gular distances determined with all the precision which modern 
mechanism has given the theodolite. The sides of the great 
imangles thus obtained being calculated, the signals form them- 
Selves fixed points for new angular measurements, and so 
the triangulation stretches from hill to hill till the prominent 
points of the entire coast are determined. The signals them- 
selves are of no small interest. Bright tin cones mounted upon 
Poles are often used, and reflect to a distance a brilliant line 
and from the extremities of this we may work backwards, so as 
