Vol. 6, 1920 
CHEMISTRY: J. LOEB 
211 
Guiana, a normal speed of 7.6 centimeters a second is recorded/ and on a 
later occasion as speed of 15.2 centimeters a second.^ 
4. There appears to be little, if any, real difference in speed towards and 
away from the nest. Large and small workers have practically the same 
speed for all temperatures during the summer months; after two months 
of low temperature, the large workers are conspicuously more active than 
the small workers (ninth column of table I). 
5. This species of ant runs as numerously in the night as during day- 
light hours; maximum activity falls between noon and midnight, accord- 
ing to the scarcely sufficient observations summarized in table IV. With- 
in the limits of 14° and 38° centigrade, temperature appears to have little 
effect on the number of ants running in the files. 
^ Wheeler, W. M., Proc. Amer. Acad. Arts and Sci., Boston, 51, 1915 (255-286), p. 259 
2 Mayr, G., Verh. Zool. Bot. Gesell. Wien., 5, 1855 (273-478), p. 319. 
3 Wheeler, W. M., Proc. Amer. Acad. Arts and Sci., Boston, 52, 1917 (457-569). 
4 Wheeler, W. M., Ihid., p. 521; and Bull. Amer. Mus. Nat. Hi '., New York, 2!, 
1905 (321-333). 
^ Woodworth, W., Univ. of Cal. Pub., Berkeley, Cal., Agr. Exp. Sta. Bull, 
207, 1910 (53-82). 
® Shapley, Harlow, Bull. Ecol. Soc. Amer., Tuscon, Ariz., 3, 1919, No. 4. 
7 Beebe, William, Jungle Peace, New York, 1919, p. 229. 
^ Beebe, William, Atlantic Monthly, Boston, October 1919 (454^64), p. 458. 
THE INFLUENCE OF IONS ON THE OSMOTIC PRESSURE OF 
SOLUTIONS 
By Jacques Loeb 
The RocKEFEiyivER Institute for Medical Research, New York 
Communicated February 21, 1920 
I. The osmotic pressure of solutions is usually measured indirectly, 
namely, by the determination of the freezing-point. Direct measure- 
ments have thus far been rare for the reason that it is difficult to prepare 
membranes permeable for water but not for the solute. The measurements 
thus far obtained on non-electrolytes by Morse, ^ and by the Earl of Berke- 
ley^ and their collaborators show that the actual values observed are not 
far from those expected according to van't Hoff's theory. 
The difficulties in the preparation of semipermeable membranes are 
considerably diminished when we substitute colloidal solutions for solu- 
tions of crystalloids. By using solutions of proteins, collodion membranes 
satisfy the demand of semipermeability. 
Gelatin is a very convenient protein for the purpose of such experi- 
ments. Like proteins in general, it is an amphoteric electrolyte being 
capable of forming salts with acids as well as with bases. The hydrogen 
ion concentration at which gelatin is neither in combination with acid nor 
with base is 10~^-W {pH = 4.7 in Sorensen's logarithmic symbol). This 
