W. O. Fenn 291 
two monovalent ions. This indicates that one ion of MnSQ, 
inhibits the dispersing effect of the other on gelatin. 
An examination of the curves for Na,SOQ., (NH,)sSO., ammo- 
nium tartrate, and sodium citrate in Figs. 4 and 5 shows that the 
valence of the anion is also important but that in high concen- 
trations the “salting out’’ effect of these salts is predominant. 
The sharp drop of these curves to the point where they precipi- 
tate the gelatin without alcohol is due to their strong dehy- 
drating powers. In the case of these salts there are two dis- 
cc. ALC. TO PPT. 
CaCle 
10 (NH4)2So, KSCN 
ae) 2 ¥ 6 
Fig. 5. Alcohol numbers are plotted against concentrations of the salts 
added to the gelatin as in Fig. 2. This figure shows how distinctly salts 
with one bivalent ion are distinguished from the monovalent salts in 
their dispersing action. 
tinct end-points to be found in precipitating with alcoho! in the 
higher concentrations. There appears first a stringy precipitate 
which is followed later by a white homogeneous one. In the 
case of Na,SO, the stringy precipitate was so opaque that it had 
to be taken as the end-point, while in the case of (NH3).SO, it 
was possible to titrate through to the second precipitate, which 
had the more usual appearance. This accounts for the fact that 
the Na.SO, curve drops after the maximum while the (NH,).SO, 
curve does not. The first white stringy precipitate is probably 
due to the ‘‘salting out’ or dehydrating effect of the SO, ion. 
The second white homogeneous precipitate is regarded as the true 
end-point. 
THE JOURNAL OF BIOLOGICAL CHEMISTRY, VOL. XXXIII, NO. 2 
