470 Philippine Journal of Science 1920 
— was made possible in these studies by the adoption of a 
new method of dealing with salt proportions. Whereas Tot- 
tingham, Shive, and others have based their salt proportions 
on the partial osmotic values of the various component salts 
(thus involving the principle of increased dissociation with 
greater dilution), the salt proportions of the present studies 
were considered without regard to ionization and were stated 
simply as molecular salt proportions. All stock culture solutions 
for any given type were planned to contain the same number 
of salt molecules per unit of volume; they differed from one 
another only in the manner in which this number was appor- 
tioned among the three or four different kinds of molecules 
(salts). When this method is adopted the several salts are 
placed together in the stock culture solution, with a certain set 
of molecular salt proportions, and the addition of water to 
this solution does not alter these proportions; they remain the 
same for all dilutions of this stock solution, although the total 
concentration is changed by the dilution process. This method 
of treatment does not at all imply that the various molecules 
of salt all remain intact in the mixed solution; it is of course 
appreciated that dissociation must be supposed to occur, and to 
a higher degree with successively lower total concentrations; 
but the degree of dissociation of each of the component salts 
in any culture solution becomes a matter for special study and 
does not enter into the simple description of the solution. 
It can be seen that this plan makes it possible to state the 
salt proportions with much greater simplicity than can be at- 
tained when osmotic salt proportions are employed; that the 
statement of atomic proportions (and also of ionic proportions, 
calling an atomic group an ion whether or not it is dissociated 
from its molecule) is similarly simplified; and that any total 
osmotic value (total concentration) that is lower than that of 
the stock culture solution in question may be obtained for any 
given set of molecular salt proportions by simple dilution from 
the corresponding stock culture solution. Of course the osmotic 
value of any solution is determined, for any temperature, etc., 
by (1) the salts employed, (2) their molecular pro 4 )ortions, and 
(3) their various degrees of ionization. This value may be 
approximately calculated by making certain assumptions, but 
the most satisfactory way to determine it is by means of the 
freezing-point method. All osmotic values mentioned in this 
paper are merely rough approximations derived from the values 
obtained by Tottingham and Shive for somewhat similar mix- 
