86 BULLETIN 1233, U. S. DEPARTMENT OF AGRICULTURE. 



during the period of its operation, because of the impenetrable soil 

 encountered below a depth of 2 feet. Little use has been found for 

 the temperature records for a depth of 2 feet, but they 'will be men- 

 tioned where they appear to give additional information of value. 



For the most part, soil thermometers have been placed in iron 

 pipes. Standard 1-inch galvanized pipe has always been used, the 

 lower end being sealed either by a threaded cap or by welding. 

 The practical advantages of this arrangement over any other type 

 of tube, or over long-stemmed thermometers which may be read 

 without moving, are obvious — namely, durability, dryness, conven- 

 ience, and cheapness. In fact, temperatures at a depth of 4 feet 

 can hardly be measured in any other way. The exposed end of the 

 pipe is capped, the suspending cord for the thermometer being 

 sealed to the inside of the cap. The standard Weather Bureau 

 "mercurial thermometer," Fahrenheit scale, with cylindrical bulb, 

 has been used mainly. To prevent any immediate change when the 

 thermometer is raised for reading, the bulb is inserted in a cork or 

 a vial of alcohol. The cork is a good non-conductor and serves as 

 a cushion when the thermometer is placed at the bottom of the 

 pipe. There is no doubt as to the accuracy of the readings at 4 

 feet when obtained in this way, since, with the bulb protected, no 

 change in the mercury is noted for 20 to 30 seconds, as a general 

 rule, while the reading may be taken in 3 to 5 seconds after removal 

 of the thermometer from its seat. 



It is, however, greatly to be regretted that in the soil-temperature 

 observations at a depth of 1 foot iron pipes have been used, since 

 at this depth these introduce an avoidable error into the tempera- 

 tures recorded. The technical problems encountered in measuring 

 soil temperatures have been discussed in detail in Eesearch Methods 

 . but for the sake of clarity it is desirable to explain the influence 

 of the iron pipe on the temperature records here presented. The 

 pipe is a good conductor of heat, and, especially if the exposed 

 portion becomes insolated, may conduct heat to a depth of 1 foot 

 more rapidly than would the soil itself. Likewise, at night, the pipe 

 may carry away heat to the air which is cooler than the soil. The 

 soil thermometer in an iron pipe, therefore, goes through a greater 

 daily range than the soil itself. For the weU-insolated control sta- 

 tion, in September, at a depth of 1 foot, the range in the iron pipe 

 found to be 7.93°, as compared with 3.41° m a wooden tube, 

 in which the temperatures always corresponded closely to those of 

 the soil itself. 



It is not admitted, however, thai the use of iron pipes has any 

 appreciable influence on temperatures recorded at a depth of 4 feet; 

 in view of the very gradual changes in such temperatures and the 

 ace of appreciable daily oscillations, it is evident that the usual 

 morning observations rive a satisfactory basis for calculating mean 

 temperatures. Also, this influence of the iron pipe is probably, in 

 the long run, completely balanced; that is, the mean temperature 

 in the iron pipe would he the same as in the soil proper, because 

 absorption and radiation by the pipe must be just about equal. 

 The pipe would, however, raise the temperatures in spring and sum- 

 mer, and Lower them in fall and winter- the general cooling period. 

 ;i this is relatively unimportant, compared with its effect on the 

 single daily temperature recorded. The great number of soil-tem- 

 perature observations in (his study have been made between 8 and 



