22 OLAF F. SUNDT 
E6étvés unit or less. There is no abnormal deviation from a mean 
gradient value for the location, measured with standard Z-type Bam- 
berg instruments. 
Pendulum apparatuses were first introduced by the Gulf Produc- 
tion Company in 1931. The company had instruments working in 
East Texas and the upper Gulf Coast to check a network of torsion- 
balance stations extending from Rains County in northeast Texas to 
Orange County and farther west. These Askania pendulums were 
contracted to the Gulf Production Company by Ludger Mintrop. 
Pendulum stations were placed at intersections of converging torsion- 
balance traverses several miles apart where possible. As a rule fairly 
good checks were obtained with not enough error seriously to in- 
fluence any possible interesting anomalies. Occasionally, however, 
one pendulum station would be several millidynes off, although con- 
verging torsion-balance traverses from other reliable pendulum stations 
would check within a millidyne or less, all radiating from fixed points 
toward that hub of divergence. When repeated, the pendulum station 
would generally check satisfactorily. 
Over long distances (80 to 100 miles), a certain gradually increas- 
ing divergence between torsion-balance and pendulum results was 
observed. The torsion balance showed a larger total number of milli- 
dynes over that distance than the pendulum. This divergence in- 
creases gradually on departing from the base station, but only in the 
direction of increase of gravity. It is proportional to the anomaly be- 
tween the first and last station of an 80-mile line. Along strike, differ- 
ences were small. : 
After study of all possible causes of this error: (1) normal values 
of torsion-balance stations, (2) latitude correction, (3) corrections 
of elevation, and (4) Bouguer correction on the pendulum, it was 
decided that none of these was consistent enough to cause such 
differences. It was finally concluded that the cause was to be found 
in the then prevailing plan of two day stations and one night station 
per large Bamberg instrument. The night-station gradients with 
‘falling temperature were always smaller in proportion to day-station 
gradients made with rising temperature. Since there were two of these 
to one night station, the tendency would naturally be to increase the 
total anomaly in an area of uniform gravity increase. These results 
were not confirmed in areas worked with small Bamberg instruments, 
or where two night and two day stations were made. 
In 1932, the Gulf Production Company introduced its own pendu- 
lum apparatus. These instruments now number ro on the Gulf Coast: 
6 field instruments and 4 base instruments near the field of work or 
as far as the radio transmitting sets can be heard by the field parties. 
These instruments use quartz instead of the invar pendulums pre- 
