14 REPORT OF THE AGRICULTURAL RESEARCH INSTITUTE 



by the presence or absence of any odd piece of earth over 

 the thermometer, would seriously impair the record; that 

 indeed it would be necessary to know the distance from the 

 surface to the instrument to within 01" if the record was 

 to be at all correct with reference to depth below the sur- 

 face. For example, if two thermometers were placed at a 

 nominal depth of 2" below the surface, and owing to rough 

 cultivation of the soil above them, the actual depth or thick- 

 ness of soil were altered to 1-5" and 2-5" respect- 

 ively, these two instruments instead of registering the 

 same temperature, say at midday, would actually show a 

 difference of more than 1° C. The rough cultivation com- 

 monly employed throughout India in the hot weather, 

 which has been styled tw hot weather weathering," implies 

 a surface much more irregular than that just exemplified, 

 and it was almost immediately clear that any record which 

 was maintained below a roughly cultivated surface would 

 be valueless because of the impossibility to define the dis- 

 tance from the surface to the instrument. The records 

 maintained at Pusa have referred to the soil below (i) a 

 smooth oare-fallow surface, and (ii) a similar surface on 

 which crops were growing. The instruments employed 

 were self-registering mercurial and alcohol thermometers 

 of a good make, the errors being determined periodically. 

 They were placed in tubes running horizontally below the 

 soil surface at 1", 2", 3", 6", 9", 12", 18", and 24" deep. 

 This arrangement is not what has been usual when observ- 

 ing soil temperatures, instruments being commonly insert- 

 ed vertically into the soil. The latter is open to criticism 

 in several respects ; the instruments are exposed to the sun , 

 they will generally possess a different, and often greater 

 conductivity than the soil; the thermometer bulbs or sensi- 

 tive part of a pyrometer indicates the mean temperature of 

 the soil stratum in which it rests, and not that of a plain 

 which is actually what is desired; if then the instrument 

 is placed horizontally, this soil stratum is considerably 

 thinner than if it is placed vertically in the soil, that is, 

 this source of error is reduced. 



