322 H. W. Turpin 
by an increase in carbon dioxide. The same fact had been previousl 
noted by Moller (quoted by E. Wollny, 1880a),by Dehérain and Demouss 
(1896), by Stoklasa and Ernest (1905), and by Leather (1915), and wa 
later mentioned by Potter and Snyder (1916). 
Carbon-dioxide production was found by the Rothamsted investigators 
(Russell and Appleyard, 1917) to be correlated with moisture and rain- 
fall. Previously E. Wollny (1880a) had observed that increasing amounts 
of water up to 9 per cent, in a quartz sand mixed with peat, resulted in an 
increase in the carbon dioxide. Dehérain and Demoussy (1896) found that 
there was an optimum water content for carbon-dioxide production in a 
garden soil. Van Suchtelen (1910) found the greatest amount of carbon 
dioxide when the soil with which he worked was 75 per cent saturated 
with water. 
The relationship observed by Russell and Appleyard (1917) between 
the rainfall of the preceding week and the carbon-dioxide content of the 
soil, was believed by them to be due largely to the oxygen dissolved in 
the rain water. That this may be true is shown by the earlier work of 
EK. Wollny (1897), and also by that of Fodor (1875), who showed that 
there is a relationship between the carbon-dioxide content and the oxygen 
of the soil, indicating that-the carbon dioxide is probably produced by 
oxidation processes. 
The crop 
The evidence available thus seems to point to bacteria as the chief 
source of soil carbon-dioxide. There are some data, however, which 
show that plants may play a considerable part in the production of this 
gas in the soil. 
Stoklasa and Ernest (1909) and Aberson (1910) noted that the roots of 
plants excrete large amounts of carbon dioxide. That the gas so formed 
is not insignificant is proved by the fact that field studies conducted 
at Rothamsted by Russell and Appleyard (1917) showed a considerably 
higher content of carbon dioxide in cropped soil than appeared in the 
bare soil, this being especially marked in May, at the time of the most 
active growth of the plant, and at the time of ripening. The same condi- 
tion was observed by Bizzell and Lyon (1918) in the case of an oat crop on 
Dunkirk clay loam, where the greatest production of carbon dioxide took 
place at about the time of blooming. Potter and Snyder (1916) observed 
