42 BOTANICAL GAZETTE [JANUARY 
28-29° C., the culture solutions were transferred to Jena distillation 
flasks and distilled after adding a slight excess of magnesium oxid. 
The distillate (about 150°¢ in bulk) was titrated against tenth normal 
acid, cochineal being used as the indicator, and the amount of ammonia 
nitrogen formed thus determined Sterile controls were run on all 
determinations to allow of an accurate estimation of the ammonia 
actually formed by B. subtilis during the period of incubation. 
Series I. CaCl, versus MgCl, 
Solutions of calcium chlorid and magnesium chlorid of 0.35% 
each were made up in bulk to contain 0.75 per cent. peptone. The 
culture solutions were made and placed in 250°° Erlenmeyer flasks 
as follows: Flask A contained 100°° of the MgCl, solution made up 
as noted above; flask B, roo*° MgCl,+5°° CaCl,; flask C, roo 
MgCl, +10°° CaCl,; flask D, 100°° MgCl, +25°¢ CaCl,; flask &, 
too°¢ MgCl,+50° CaCl; flask F, 100°° MgCl,+100°° CaCl. 
Then beginning at the other end of the series, flask K contained 1oo°* 
of the CaCl, solution; flask J, roo®* CaCl,+5°* MgCl; flask J, 
too’* CaCl,+10°° MgCl; flask H, roo CaCl,+25°° MgCl; 
flask G, too** CaCl,+50°° MgCl. 
After the solutions were made up, the mixed solutions in each 
flask amounted to more than too°°, while the flasks with pure salts 
(plus peptone) contained only 1oo°°. In order to make the con- 
ditions the same for all the flasks (particularly in respect to the 
amount of fluid surface exposed to the air), enough liquid was pipetted 
out of the flasks containing mixed solutions to leave just 100°° in 
each flask. The flasks were then sterilized in the autoclave at 1.25 
atmospheres of pressure for 30 minutes. A 48-hour culture of B. 
subtilis in 1 per cent. peptone was used as inoculating material. By 
slight shaking the membrane formed at the surface was precipitated 
to the bottom, and by tilting the flask to one side and carefully setting 
it down again, one part of the bottom of the flask remained free of 
membranous material and the liquid above was homogeneous in 
character. Of this homogeneous liquid 1°° was taken for inoculation 
in each of the flasks prepared as above described. The table of 
results follows: 
