BELLE ISLE TORSION-BALANCE SURVEY 1347 
TABLE III 
VERIFICATION OF PREDICTED DEPTH AND THICKNESS OF THE CAP ROCK 
(Wells of Freeport Sulphur Company) 
Depth to Top of Cap| Thickness of Cap 
in Feet 






Well Cor Bead EB Rea: Total | Character of Cap 
Number Pre- |Actual| Error | Pre- |Actual| Error Depth Rock 
dicted dicted 
I 850 | 1,247 |—397 470 | 442 |+ 28 | 1,853 |Hard lime rock with cal- 
cite veins, mostly im- 
pervious 
2 690 ; 660 |+ 30] 510 | 644 |—134 | 1,321 |Same as foregoing 
3 1,180 | 1,672 |—492 | 410] 253 |+157 | 1,940 |Same as foregoing 
4 550 | 303 |+157 | 470] 174 |+296 | 828 |Limestone and some gyp- 
sum and anhydrite 
5 650 | 546 |+104 | 500]! 327 |+173 { 880 |Same as foregoing 
6 860 | 1,058 |—198 | 740 | 387 |+353 | 1,445 |Broken lime rock and cal- 
cite; voids filled with 
sandy shale 
7 620 | 426 |+194 | 730! 149 |-+581 | 580 |Broken lime rock, calcite 
and gypsum; voids 
filled with sandy shale 
Percentage errors in predicted depth to the top of the cap: 
46, 4, 41, 28, 16, 23, 31 Numerical mean, 27 per cent 
Percentage error in predicted thickness of the cap: 
6, 26, 38, 63, 34, 47, 79 Numerical mean, 42 per cent 
CAUSES OF ERRORS 
The errors in the calculated predictions probably arise from a series 
of causes. 
The unfavorable conditions under which the observations were made 
are sufficient to account for considerable error in the magnitude of the 
observed gradient values. Because of the marsh terrane, it was imprac- 
ticable to maintain our normally ngorous limits of error in the readings. 
The impracticability of taking the terrane levels and of making the ter- 
rane corrections probably led to the introduction of some error into the 
observations. An error of even 1 E with the same sign at three con- 
secutive stations critically placed at the edge of the dome easily might 
cause considerable errors in the calculated position of the cap and in the 
calculated depth to the top of the cap on the flank. The error shown by 
well No. 1 on the west radius could be caused by an error of +2 E in the 
observed values of the gradient in the two stations above the predicted 
edge of the cap on that west radius. Errors in the observations at those 
stations are especially probable as the stations are near a bayou. 
65 
