CHEMICAL AND BACTERIOL°DGICAL STUDY OF WHEAT SOIL. 99 
table this fact comes out. sharply—that during winter and 
spring months there is a remarkably low average of nitrates 
present in the soil. It is true there is a spasmodic rise in 
four cases in September, but it was not till the end of the 
year that the real increase came. It must be remembered 
that the active growth of the crop would keep certain plots 
in check, but what can be said of the meagre increase in 
the cultivated fallow? The most striking feature is the 
sustained increase beginning on all soils between November 
and January, after the wheat crop is normally ripe. 
*“Lucerne’’ plot shows up in a very interesting light, 
‘because of its low moisture content and different constitu- 
tion. 
It is not easy to compare other countries in respect of 
nitrate cycles for a year for the reason that weather con- 
ditions, soil composition and state, previous treatment and 
the impossibility of theorising on anything but an extended 
period of investigation raise a strong barrier to such com- 
parisons. There are of course variations in nitrifying 
power in relatively small areas, and other workers even 
under similar climatic conditions are dealing with vastly 
different soils. ‘The soil used is fairly typical of large areas 
on the South West Slopes, and that makes the comparison 
the more extraordinary. Hallin South Africa has published 
a list of monthly determinations for an average soil at the 
Potschefstroom Station which runs as follows:— 
Parts per Million of Nitric Nitrogen. 

| | Jusel July Was | Oct. | Nov Dee. 
= | : 
Cultivated Fallow!34'8 24-0 16-0 26:0 35-0 |49-0| 38 | 13* 
| Teff | 88) 87) F3111-1/ 88 li46! 20] s* 
* Average of two determinations per month. 

Jan, | Feb. | Mar. | Apr. | May 
7) 819" 
14-54] 6-4 | 6:5 | 











He reports further his highest figure for nitrate was 55 
p. pm. at midsummer, when it may be added maximum 
temperature and precipitation give ideal conditions for the 
