APPLICATIONS OF GEOTHERMICS TO GEOLOGY © 31 
to a depth of about 300 feet. A similar condition exists at Birming- 
ham, Alabama. Barring the possibility of abnormal conditions in the 
well, which seems rather improbable, the flattening of the curves may 
be due to ground water. 
In summarizing the remaining evidence of the preceding sections, 
it can be said that it tends to show that regional variations exist; 
and that local variations have been found on five different types of 
structure on which oil fields are located. They are: faults, salt domes, 
granite ridges (or simple types of unconformity), sand lenses, and a 
structure with a large closure in which the temperature may be con- 
trolled in part by the movement of water. Elsewhere (31), the writer 
has discussed the evidence in regard to the variation of temperature 
over structures of both large and small closure in which the move- 
ment of water is probably of no importance. 
With regard to regional variations, an hypothesis of vertical up- 
lift proposed by Nevin and Sherrill (32) to explain the origin of certain 
local uplifts in north-central Oklahoma is of fundamental importance. 
According to their hypothesis, regional tilting of a rock mass by ver- 
tical forces has resulted in producing local uplifts over areas of weak- 
ness in the pre-Cambrian complex. By referring to Figure 6, it will 
be noted that the temperatures rise gradually as the depth to the 
granite diminishes, as we pass from Oklahoma City to Wewoka, and, 
furthermore, the highest temperatures are found over the oil fields, 
the areas in which, according to Nevin and Sherrill, the hot rocks 
have been pushed upward through the basement complex. The proc- 
esses of uplift thus proposed by Nevin and Sherrill suffice to explain 
both local and regional variations of temperature in certain areas. 
Another important contribution to our hypothesis of regional varia- 
tions was made a number of years ago by N. H. Darton (33). His map 
shows clearly that the temperature gradients in eastern South Dakota 
increase as the depth to the granite or quartzite diminishes. As shown 
on the map (Fig. 5), the reciprocal gradients vary from 1° F. in about 
20 feet to 1° F. in more than 50 feet. 
In order to reach a conclusion in regard to other causes of local 
variations, let us consider the possibilities of thermal conductivity, 
chemical reactions, and radioactivity. 
First, let us consider the differences in the values of 1/b at Haver- 
hill, Kansas (Fig. 10), from the standpoint of thermal conductivity 
and the generation of heat from chemical reactions. 
The sand lens can be regarded as a buried disk that is a very poor 
conductor of heat, for, under the same conditions, water transmits 4 
times and limestone 14 times as much heat as petroleum. The correct 
593 
