CHEMICAL PROPERTIES OF HAY INFUSIONS 279 
the sixth day a sample taken from just above the hay in infusion 
C-31" titrated 20, against 136 an inch below (in the midst of 
the hay), and 10 at the top of the infusion. It is thus clear why 
the tops of typical infusions of series C are so tardy in reaching 
their maximum acidity (fig. 1). 
Infusion M-51” on the second day of its history was showing 
an acidity difference of 41 between the top and bottom. It was 
then thoroughly stirred after which samples from both top and 
bottom titrated 20. By the next day an acidity difference had 
again developed to the extent of 20. It was again thoroughly 
mixed (both top and bottom then titrating 29), and two days later, 
an acidity difference (to the extent of 9) was once more apparent. 
From here on, however, equilibrium was soon reached. This 
further illustrates the point made that acidity does not increase 
uniformly with the depth. On the second day of this infusion 
the top and bottom titrated 15 and 56 respectively. When thor- 
oughly mixed the infusion gave a uniform titration of 20 instead 
of the average 35. 
In the early history of the typical infusion of series C, the liquid 
in the immediate region of the hay is highly colored, while the 
liquid above is almost colorless. The approach to equality in 
acidity can be followed roughly by the gradual diffusion of this 
colored material. | 
The observations on acidity differences may be summarized 
as follows: 
1. The greatest acidity is found near the source of supply of 
soluble material upon which the bacteria can act. 
2. The difference between top and bottom acidity depends 
upon the relative concentrations of acid-yielding hay constitu- 
ents in these regions. 
3. When there is uniform concentration, or when diffusion keeps 
pace with solution, there is no essential difference between top 
and bottom acidity. 
From table 5 we see that the curve for ‘methyl-orange alkalin- 
ity’ follows an irregular course with a general upward tendency, 
as also noted by Peters. The latter’s figures for these determina- 
tions are from three to ten times as great as here recorded. This 
