126 ' E. E. ROSAIRE AND JOSEPH L. ADLER 
profile a computer can not plot the dips to so great a degree of accu- 
racy. Indeed, the difference between the dips of the several reflections 
obtained at different horizons from the same shot is usually several 
times this amount. In drawing the phantom horizon, however, the 
geophysicist, realizing which dips conform best to the structural pic- 
ture he is making, will naturally select those yielding the least mis- 
closure. This personal element doubtless accounts to some extent for 
the surprisingly small value of the probable error; but it must also 
be remembered that the error here determined takes into account the 
compensatory effect of numerous observations. 
Another considerable error may enter into the computation of the 
dips through incorrect time determinations. Such errors may be due 
either to phase lag in the recording instrument, or to the personal 
element in picking the exact time on the record. Errors of more than 
0.002 second due to these two causes are probably not unusual,® and 
might produce errors in the time difference of at least 0.0015 second. 
While such an error in determining the time step-out would produce 
an error in dip of more than 1°, the compensatory effect of the numer- 
ous observations made in any traverse will reduce the net error to a 
fraction of this amount.® 
Other sources of error dependent upon inaccuracies of measure- 
ment lie in the determination of the thickness of the low velocity zone 
under the first and last geophones, computations based on an incor- 
rect velocity for that zone, and use of an incorrect velocity in the 
underlying high-velocity zone. Of these three sources of error, the 
first is the most serious. Where great variability in the thickness of the 
low velocity zone exists, its relative thickness under each of the end 
seismometers must be determined as accurately as possible before any 
validity can be attached to the dips computed. A difference of 3.5 
feet in the thickness of the low-velocity zone between the ends of the 
geophone spread may easily produce an error of about 1° in the com- 
puted dip if not properly corrected. 
Errors in the velocity itself are less important. While the use of an 
erroneous velocity may produce an error in dip of several degrees, 
since shot points are located at equal distances on opposite sides of the 
seismometer position, the errors will be in opposite directions in the 
pair of dips so obtained from each reflecting horizon. They thus com- 
pensate each other almost exactly. If the error is great this alterna- 
5 C. B. Bazzoni, personal communication. 
® Paul Weaver, paper read before the Houston Geological Society, September 21, 
1933. 
648 
