27 2 DIFFUSION, OSMOSIS, AND FILTRATION. 
maintained that cessation of motion of bacteria placed in solutions 
indicates that the solution is isosmotic with the cell sap, in cases 
where poisonous action can be excluded. 
The above physiological methods of measuring osmotic pressure are 
of considerable interest, but, as already stated, their application is 
decidedly limited: and though, as will appear below, often of indirect 
value, as giving information bearing on the permeability of cells and 
membranes, they are not to be classed with the more accurate methods 
of experimental physics. 
In the case of colloid solutions, it is not necessary to use pre- 
cipitation membranes for the direct measurement of osmotic pressure, 
for such material as vegetable parchment is, as a rule, impermeable to 
colloids, and it moreover presents certain advantages in particular 
cases, namely, those in which the colloidal substance is contaminated 
by salts {e.g. albuminous solutions), since the salts can pass out. and 
the determination is freer from the error of inclusion of the partial 
pressure of these, unavoidable by a direct measurement by a copper 
ferrocyanide membrane, or an indirect determination by lowering of 
freezing-point. 
It is known that solutions of colloids of considerable concentration 
exert very low osmotic pressure, though their exact measurement is 
difficult. Picton and Linder, 1 in a direct measurement (by a copper 
ferrocyanide membrane) of the pressure of a 4 per cent, solution 
of colloidal arsenious sulphide, obtained a pressure of only 17 mm. of 
water. Sabanejew 2 states that the lowering of freezing-point by silicic 
acid is so small as to be within the limits of the method. With 
albuminous solutions the difficulty of contaminating salts is almost 
insuperable, and since the molecular weight of albumin is not known, 
calculation is excluded. 
Sabanejew 3 investigated the lowering of freezing-point of water by 
solution of egg albumin, and cpiotes a lowering of "02° C. for a 15 - 6 per 
cent, solution, and -042° C. for a 30-35 per cent, solution, but since the 
specimens held -4 to - 66 per cent, of ash, the numbers are of no value. 
Tamman 4 gives the difference in lowering of freezing-point of horses' 
serum, produced by coagulation of the proteids by heat and removing 
them, as only -006° C, which is in the region of the error of the method. 5 
Dreser 6 and Koeppe 7 also state that the removal of proteid from 
albuminous solutions does not affect the osmotic pressure, while 
Ludeking 8 maintains that the boiling point of 40 per cent, solution 
of gelatin is 100° C. 9 It is therefore uncertain whether proteids in 
1 Journ. Chem. Soc., London. 1895, vol. lxvii. p. 63. 
2 Bcr. d. deutsch. chem. Gesellsch., Berlin, 1890, Bd. xxiii. S. 87. 
3 Ibid., 1891, Bd. xxiv. S. 558. 
4 Ztschr. f. physikal. Chem., Leipzig, 1896, Bd. xx. S. 180. 
5 Starling, on the other hand, quotes two experiments to prove that the osmotic 
pressure of the proteids of serum can be directly measured. It is stated to he from 30 to 
40 mm. of Hg. Journ. Physiol., Cambridge and London, 1895, vol. xix. p. 323. Cf. also 
next article, p. 308. 
15 Arch. f. exper. Path. it. Pharmakol., Leipzig, 1892, Bd. xxix. S. 314. 
7 Arch. f. d. ges. Physiol., Bonn, 1896, Bd. lxii. S. 571 (footnote). 
8 Ann. d. Phys. u. Chem., Leipzig, 1888, Bd. xxxv. S. 552. 
9 A lowering of vapour pressure (raising of boiling point) is produced by solution 
of a substance in a solvent, and the lowering of vapour pressure, like that of the 
freezing-point, is proportional to the concentration. Wiillner, Ann. d. Phys. u. Chem., 
Leipzig, 1858, Bd. ciii. S. 529 ; 1858, Bd. cv. S. 85 ; 1860, Bd. ex. S. 564 ; Tamman, 
Ann. d. Phys. u. Chem., Leipzig, 1888, Bd. xxxiv. S. 299. 
