EXPERIMENT STATION BULLETINS. 693 



ance bulbs and 85 feet wire leads. Tlie leads were incased in lead pipe 

 so as to keep them dry and well insulated. The ends of these leads were 

 run into a small iiouse specially constructed for the purpose. In this 

 house there was kept, on a permanent stand, a balance indicator upon 

 which the temperature was read. 



During the first year, from Dec. 1, 1911 to Nov. 30, 1912, the surface 

 of all the five different types of soil was covered with a very thin layer of 

 sand in order to eliminate their differential factor of color. The tem- 

 perature during this period was also measured by an indicator graduated 

 to only 1°F. 



During the second year, from December 1, 1912 to November 30, 1913, 

 the thin layer of sand was removed from the surface of all the soils and 

 consequently their natural surface was exposed to the atmosphere and 

 allowed to play its role. Beginning with this year also the old balance 

 indicator was replaced by a new one w^hich was graduated to 0.5°F and 

 thus more sensitive than the first. 



In the third year, from December 1, 1913 to November 30, 1914, all the 

 soils were again covered by a thin layer of sand. 



At the beginning of the fourth year, however, from December 1, 1911 

 to November 30, 1915, the surfaces of all the soils were again uncovered, 

 and the thermometers were removed from the 6, 12 and 18 inch depths at 

 which they had remained from the beginning of the experiment, and 

 placed at 2, 4, and 6 inch depths, respectively. During the warm part of 

 this year the surface temperature was also taken by means of mercury 

 thermometers having extremely small bulbs. 



It was stated above that the temperature records were taken at 7 

 A. M., 12 M. and 6 P. M. These periods, how^ever, were changed some- 

 what during the year in order to obtain as true an average and range of 

 temperature as possible. During the warm part of the year, when the 

 sun rose early in the morning, the temperature records were taken at 

 about 6 A. M., 1 P. M. and 5 P. M., while during the cold part of the 

 year they were taken at about 7 A. M., 12 M. and 6 P. M. The daily 

 average temperature, therefore, is composed of these three diurnal read- 

 ings, and the daily amplitude, the difference between the highest and 

 lowest temperature recorded during the day. 



Even though the temperature was recorded three times a day it can- 

 not be claimed, however, that it gives the absolute average and range 

 of temperature for all depths and for all soils. This is at once obvious 

 from the fact that the real maximum and minimum temperatures at the 

 various depths of the same and of different soils are attained at entirely 

 different periods of the day. The three diurnal records, however, do 

 give a very close approximation of the true daily average and fluctuation 

 of temperature for all soils and for all depths. Furthermore, since these 

 three diurnal records have been taken almost every day for four years the 

 average for the month, season or year is exceedingly close, if not exactly, 

 the true average and range of temperature. 



As already mentioned the previous report^, contained all the data for 

 the first year. In the present report will be presented the results for the 

 following three years, or from December 1, 1912, to November 30, 1915. 



•Technical Bui. 17,'Mich. Expt. Sta., 1913. 



