28 BULLETIN 1059, U. S. DEPARTMENT OF AGRICULTURE. 



6. Influence of air temperatures and light upon soil temperatures, 

 especially maxima, with different kinds of soil cover. 



7. Correlation between soil temperature and extremes of drought 

 in the surface soil. It should be noted that the distinction between 

 drought injury and heat injury to young plants is often very diffi- 

 cult, as is shown by Hartley's (44) work. Also, that the soil sur- 

 face can not for long be excessively hot without becoming arid. 



Time of Observations. 



The daily range of temperatures at the surface of the soil may be 

 considerably greater than in the air above, and for the study of sur- 

 face conditions the thermograph is essential. The time of observa- 

 tion of thermometers used to check this instrument should be a time 

 when radiation and absorption of heat in the surface soil arc about 

 equal, or, in short, in the early morning. At no other time will a 

 real check be found possible, because the thermometer and thermo- 

 graph are not equally sensitive to changes and do not absorb direct 

 sunlight equally well. 



The daily range of soil temperatures at a depth of 1 foot or more 

 is so slight that it is unimportant, except in its bearing on the ques- 

 tion of determining the mean for the day. The latter must often 

 be obtained from a single daily reading of soil thermometers, and 

 must be based on a knowledge of the diurnal oscillation for the 

 particular site. The daily range at 1 foot will seldom, if ever, exceed 

 5° F., and at 2 feet is far less; so that, at greater depths than 1 foot, 

 almost any time of the day is suitable for obtaining approximately 

 a mean temperature. The time of observations may therefore be 

 made to accord with other observations without any serious disad- 

 vantages. 



The point of this discussion is that it is not satisfactory merely 

 to compare the soil temperatures of several sites for a certain time 

 of the day, since at the time one soil may be cooler than the mean 

 temperature for the day and another above the mean. 



Daily Mean Soil Temperatures. 



The simplest way to secure a proper comparison of sites in respect 

 to mean soil temperatures would, of course, be to determine the 

 maximum and minimum for each day and to average these, as is 

 commonly done with air temperatures. However, as will be pointed 

 out in the discussion of apparatus, registering thermometers for 

 this purpose have not been satisfactorily developed; so that, at 

 present, dependence for a complete record must be placed on one of 

 the several types of soil thermograph, supplemented by frequent 

 readings of a thermometer. 



