ON COLLOID CHEMISTRY AND ITS INDUSTRIAL APPLICATIONS. 133 
here. The potential difference between two liquid phases, referred 
to by Loeb, has clearly the same origin as the membrane potential of 
the cell. The former has been investigated by Haber and Klemen- 
siewicz (1909) in a well-known paper. If the salt is hydrolytically, 
as well as electrolytically, dissociated, a more complex state of affairs 
exists, which has been investigated thermodynamically by Donnan 
(1911). If, for example, we are dealing with a sodium salt of a 
weak acid, there are no forces to restrain the free diffusion of sodium 
hydroxide through the membrane, so that the alkali would be 
detected on the outer side, while the colloidal acid inside might be 
precipitated. Moreover, when no hydrolysis is present, if an acid, 
even as weak as carbonic, is present outside, the sodium ions at the 
membrane are partially replaced by hydrogen ions, the sodium com- 
bines to form carbonate and, by renewal of the outer fluid, all of the 
sodium can in time be removed. 
III. Secretion. Although it is clear that complex changes of 
permeability occur in this process, too little definite knowledge is at 
hand to make detailed discussion of profit. The reader may be 
referred to my “General Physiology,” pp. 163 and 334, for a brief 
statement. The point of immediate interest is that, suppose we have 
an inverted U-tube with a semi-permeable membrane at both ends 
and filled with a solution of cane sugar. If we immerse both ends 
in water in two separate vessels, a large osmotic pressure wil! develop 
inside the tube, but no liquid will escape, provided that the mem- 
branes can withstand the pressure. Now imagine the semi-permeable 
membrane at one end to become permeable to the sugar. A current 
of water, carrying sugar in solution, will pass through the tube from 
one vessel to the other, as long as there is any sugar left in the tube. 
This tube may be compared to the cells of a secreting gland, one end 
being in contact with lymph, filtered from the blood, the other end 
with the watery secretion in the duct. Water can be conveyed in 
the way described by a change in the permeability of the end 
bordering on-the duct. Cases of this kind have been described by 
Lepeschkin (1906) in plant mechanisms. 
IV. The Blood Vessels. Scott (1916) showed that when liquid is 
absorbed into the blood from the tissue spaces, this liquid, while 
containing all the crystalloids, is free from the colloids. The walls 
of the blood vessels are therefore impermeable to colloids, If, then, 
these colloids are such as have an osmotic pressure, the conditions 
are such that it can be manifested and it will play an important part 
in the passage of water from the blood to the tissues, and vice versa. 
A few words are necessary therefore on the osmotic pressure of 
colloids. Starling (1896) showed that those present in the blood 
have an osmotic pressure of about 30 to 40 mm. of mercury at room 
temperature. Later work by Moore and Roaf (1997), Donnan and 
Harris (1911), Sérensen (1918) and myself (1911) showed that there 
are many colloids whose active elements are sufficiently small to give 
a fairly high osmotic pressure. When the colloid is an electrolytically 
dissociated salt of a diffusible ion with one to which the membrane 
is impermeable, as in the case of congo-red with a parchment paper 
membrane, an interesting question arises, whether the diffusible ions, 
which are held only by electrostatic forces, play their part in the 
20895 E 3 
