SEISMOLOGICAL DISCOVERY OF VERMILION BAY 59 
flank, with disappointing results. Well No. 5, supposedly a location 
similar to Well No. 2, was abandoned in rock at 6,500 feet. Well No. 
6, well inside this estimated edge, was abandoned at 5,824 feet with- 
out finding dome materials, when cap or salt was expected not deeper 
than goo feet. A recent revision of the interpretation on this flank was 
made by the junior writer, and led to the location of this edge as shown 
by the dashed line. As no further wells have been drilled since No. 6, 
the original interpretation of the northwest and southwest flanks, 
and -the re-interpretation of the northeast flank remain unchecked 
to date. 
With the experience accumulated, since this detail, on several 
other domes, the writers are in a position to comment on the job. It is 
obviously incomplete, since four points, no matter how well deter- 
mined, can hardly furnish satisfactory control on the periphery of a 
salt dome like that in Vermilion Bay. Later practice was to locate 
eight points, reversing all profiles. These additional points, determined 
on stronger evidence, resulted, in the case of the Lake Barre salt 
dome, in commercial production by the first flank test, followed by a 
string of nine successive flank producers around the periphery of the 
dome. 
In addition, the profile data on the northeast flank somewhat 
definitely indicate that this flank slopes less steeply than any of the 
other three. Failure to realize the amount of this slope (as shown by 
the apparent velocities) undoubtedly played a major part in the 
erroneous location of this flank. 
Finally, these conclusions were reached on profile data with too 
low a survey density. Recorder intervals of 1,500 feet can hardly be 
expected to locate salt dome flanks to plus or minus 250 feet. In the 
later details, these recorder intervals were cut down to as little as 250 
feet, along the profile. 
In presenting the data for publication, attention is called to the 
fact that curved lines have been used to approximate the travel-time 
curves. At the time of this survey, field practice was to use success- 
sive, intersecting straight lines, but for the past 3 years, curved path 
interpretations have entirely replaced the older, empirical, strong- 
arm methods of attack in almost all phases of refraction surveys. In 
view of this, the writers feel justified in restricting their case treat- 
ment to the earlier predictions, and so indicating, to one experienced 
in the art, the contrast between the old and the new interpretation 
methods. 
389 
