however, we are on the ground and therefore not stationary but moving with the
earth about its axis. The accelerations resulting from this motion must be re-
moved by computing the centrifugal effect of the earth's rotation and inserting
it into the system. "This is easily done at the airbase since its latitude and mag-
netic declination are known. These quantities need not be known to any great
degree of accuracy (a few degrees of error being negligible) since the maximum
value of the correction does not exceed 6 minutes of arc.
Almost all mapping is done with reference to a pendulous vertical (level
bubble in a surveyor's level). It is consequently very desirable to produce a
reference which is aligned with the local plumb bob vertical, rather than with a
geocentric vertical. The alignment procedure described in the previous para-
graph would result in a geocentric vertical. At the expected position of photo-
graphy there will be a deflection of a plumb bob from the geocentric vertical due
to the centrifugal effects of the earth's rotation. If the latitude and magnetic
declination of the area to be photographed is inserted into the system so as to
add the required angles to the indicated vertical we have the required reference.
As previously noted, the correction need only be approximate, an error of one
minute resulting from a latitude error of +5 degrees or a heading error of +10 de-
grees, This corresponds to an area about 700 statute miles in the North South
direction by at least an equally great distance in the East West direction. It
is therefore felt that the required data can probably be set in as a preflight ad-
justment, although it can easily be made variable in flight.
In addition to the rotation effects previously discussed, there is an addi-
tional faulty assumption. This lies in the assumption of a spherical earth. In
fact this is not so and it affects the system because the earth's ellipticity causes
-9-