SURVEY OF MUNICH TERTIARY BASIN 75 
by one or more different formulae; (3) corrected also for isostatic com- 
pensation. In geodetic surveys, the absolute value of Z is reported for 
the date of the observation or it may be corrected to some standard 
date. The value of gravity varies with astronomic latitude and the 
value of Z with magnetic latitude. In geodetic work, commonly no 
correction is made for variation with latitude. A variation of gravity 
which can be calculated from the results of torsion-balance surveys is 
the variation of gravity at any-equipotential surface, that is, level sur- 
face which lies approximately at the general elevation of the surface of 
the area, for (1) the horizontal gradient of gravity, which the torsion- 
balance measures, varies slowly with elevation and (2) torsion-balance 
surveys are made only in areas of slight relief. That variation of 
gravity would be comparable with the variation of geodetic observed 
values of gravity reduced by the ‘free air” reduction to a common 
elevation essentially that of the surface. Torsion-balance surveys 
such as those of the B.M.I. are made in search of structure; the effect 
of the variation of gravity with latitude ranges from 6 to 8 Eétvés 
units? in temperate latitudes and therefore seriously obscures the 
very many structural anomalies whose maximum gradient may range 
only from 7 to 15 Eétvés units. The effect of the latitudinal variation 
of gravity, therefore, is eliminated in the station calculations of the 
gradient. Relative gravity which is calculated from torsion-balance 
observations, therefore, is not directly comparable with any of the 
various types of values of gravity which are given in geodetic reports. 
The vertical component Z of the earth’s magnetic field similarly varies 
with magnetic latitude; the latitudinal variation obscures the struc- 
tural anomalies and therefore is eliminated in the routine of the sta- 
tion calculations in most magnetic surveys by oil companies. The 
observed variation of Z, of such surveys, therefore, is not directly 
comparable with the variation of Z in geodetic magnetic surveys. 
The observed variation of gravity (Ag,) of this paper, therefore, 
is the horizontal variation of gravity at a common level, which is any 
level between 300 and 500 meters elevation above sea-level,—cor- 
rected for topography and for the variation with latitude. If the 
earth’s crust and interior were perfectly homogeneous, no variation 
of gravity would be observed. Any inhomogeneity in the earth’s in- 
terior could produce faint large-scale anomalies. The variation of 
density in the earth’s crust and, mostly, in the upper kilometers of 
the earth’s crust produces most of the observed anomalies of this 
paper and of surveys similar to the B.M.I. torsion-balance survey. 
36 to 8X10 * C.G:S. units per horizontal centimeter. 
597 
