96 J. K. TAYLOR. 
Bacterial Numbers. 
Bacterial counts were made on standard agar neutral to 
bromthymol-blue. It*was decided to continue with the 
medium since it is a very fair general one, though synthetic 
agars usually give higher counts. Its drawback was the 
constant occurrence of spreading colonies which were 
capable of overrunning a plate in 48 hours. Since a six- 
day period was made the basis of the counts serious losses 
occurred and caused gaps in the table of results. Table IIT. 
gives a skeleton picture of the variations in numbers. 
The numbers are probably averages for such a soil and are 
roughly comparable with other semi-arid soils. 
Table I11.—Bacterial Count from June 1921 to March 1922 
(in millions per gramme ). 





Date of Sampling. | Grass | Lucerne PA PB PC 
Jane 24, 1921 we 2:0 a PA, Bb; C— Go 
Sealy la eo ersnueed ee 3.75 | O75 | 3-75 
ae Ma ls || UDG 3-0 4+] 5:1 
Sep. 15 2:65 | 2-71. |, 9-4 Son anieeD 
Oct. 13: 2. P| liom oe! 5-9 55 ion 
Nowa) 162. ey meee 3°05 eo! See 
Dec lpia 2°58 3°2 6:0* foe oO 
Feb. 15, 1922 2:25 | 5-0 aed 6-0 6-0 
Nar se liG mer: 3°] 5:6 On0k 71 7°63 




* Circum. 
The following points are worth emphasising (1) The order 
of merit of the plots is fairly constantly as follows :— 
cultivated fallow, cropped plot, uncultivated fallow, grass- 
land. (2) The drying of the soil does not seem to vitally 
affect the total numbers and in summer they are consider- 
ably higher than in winter. (3) There was a distinct drop 
in the bacterial numbers between July and September. 
The soils from PA, PB; PC and probably from “‘ Lucerne’’ 


