From Table 5 it will be seen that if the previously found estimate 

 for the diffusivity is approximately correct, the lag between the 3 in. 

 depth and the 5 in. depth should be in the vicinity of 3-3 hours. It 

 thus appears that heat is conducted downwards between the 3 in. and 

 5 in. levels at a faster rate than is consistent with the normal theory of 

 conduction of heat ■ in a solid in as much as the methods depending 

 respectively on the amount of lag and the decrease of daily range between 

 the two levels do not, in this case, yield concordant results. The lag of 

 1-5 hours corresponds to a value of 6-1 x 10~^ C.G.S. units, which is 

 in reasonable agreement with the diffusivity found by Johnson and 

 Davies for chalky soil on the Salisbury Plain. Johnson and Davies 

 found no disagreement between the results of the lag method and the 

 ratio method, but results similar to those found in this paper have been 

 emphasized by Wright, who attributed the discrepancy to the presence 

 in the top layers of the soil of numerous grass-roots which somehow 

 interfere with the normal process of conduction. 



Part II — Observations in Other Localities 

 Wairenga-o-kuri, Gishorne District 



In view of the anomalous transfer of heat that has been found to 

 take place in the surface layers of the soil at Earnscleugh it was thought 

 worthwhile to make a brief examination of earth thermograph records 

 from other parts of New Zealand in order to see if any similar phenomenon 

 occurred. The most convenient set of records is that from Wairenga- 

 o-kuri, in the Gisborne district, which give continuous records of tem- 

 perature at 4 in. and 8 in. below the surface of bare soil. The results 

 obtained at Wairenga-o-kuri for a period of observations during October, 

 1947, are shown in the self-explanatory Table 6. 



This table shows that the mean lag is about 3-5 hours, and for a 

 difference of depth of 4 in. we find that this corresponds to an average 

 value of 4-5 X 10~^ C.G.S. units for K. The mean ratio of the ranges 

 observed at 4 in. and 8 in. depth is 2-56, which corresponds to a value 

 of K equal to 4-3 x 10~^. Thus in the soil at Wairenga-o-kuri the values 

 of K obtained by both the ratio method and the lag method are perfectly 

 consistent and lead to a ver\' reasonable value for the diffusivity. Here 

 there is no question of anomalous transfer of heat in the soil. Records 

 from other parts of New Zealand have also been examined and tend 

 to show that the transfer of heat is, in general, normal. 



Table 1 



54 



