USE OF GEOELECTRIC METHODS IN SEARCH FOR OIL 175 
were found on a few occasions. It is when these give rise to negative 
values in the calculated resistance that they are especially conspicu- 
ous. With some types of apparatus which have been used in resistivity 
surveys, such absurd results are not forced upon the attention of the 
observer, but the error is not less serious for his being unaware of it. 
The errors which may arise from defective insulation on the field 
c2bles are nearly always additive. It is only for some unusual positions 
of the cables that they cause the measured values to be too small. 
However, in all our experience, errors from this source have been 
inappreciable except on a few occasions when old cables were used in 
rainy weather. 
It is in the instruments that troublesome insulation defects may 
develop. The errors which result from these are in part dependent 
upon the relationships between the circuits in the instruments and 
upon the contact resistance of the electrodes. Since both these factors 
may be varied to some extent, this provides a means of detecting such 
errors. Four different circuit relationships are readily obtained by 
changing the connections of the field cables with the instruments. 
The values of resistivity obtained from measurements with these 
different combinations may all differ if leakage effects are present. 
‘Their mean is in general not free from leakage error as has been as- 
sumed by some. These errors are completely eliminated from this 
mean only when the contact resistances of the pick-up electrodes or 
those of the energizing electrodes are equal, and this holds only if the 
condition of the insulation remains constant during the series of 
measurements. 
The uncertainty and inconvenience which attend attempts to 
eliminate these errors by such means can, however, be avoided by 
making suitable provisions in the apparatus. The device which has 
been adopted for this purpose is similar in conception to the “guard 
ring’”’ which is often used, especially in instruments employed for 
measuring very small quantities of electricity. This consists essentially 
of an independent system of conductors so arranged as to intercept 
all possible paths by which current may leak from the energizing 
circuit to the pick-up circuit. If this guard system is connected to 
earth at a relatively neutral point, its potential is maintained at a 
value which differs so little from that of the pick-up circuit that this 
serves to reduce such errors to several orders’ lower magnitude than 
those which would result if this device were not employed. In thou- 
sands of tests made by W. J. Rooney under a wide variety of con- 
505 
