OSMOSIS. 309 
transudation, until we get equilibrium established at a somewhat higher 
point, when there is a mere dilute fluid in the tissue spaces, and there- 
fore a higher absorbing force to balance the increased capillary pressure. 
With diminished capillary pressure there will be an osmotic absorption 
of salt solution from the extra vascular fluid until this becomes richer in 
proteids, and the difference between its osmotic pressure and that of the 
intravascular plasma is equal to the diminished capillary pressure. 1 
Here, then, we have the balance of forces necessary to explain the 
accurate regulation of the quantity of circulating blood according to the 
conditions under which the animal may be placed, and it seems 
unnecessary to invoke the aid of vital activity to explain the process. 
Certain corollaries of this mode of explanation agree well with observed 
facts of experiment. Thus the more impermeable the capillary the 
smaller will be the amount of proteid exuded with the lymph. A 
higher capillar}- pressure will therefore be needed in its production, and 
there will be an equally high force tending to its reabsorption. A rise of 
capillary pressure will only increase the amount of lymph in the 
extravascular spaces to a certain extent, but will at the same time cause 
this lymph to be more dilute, so that there will be a corresponding rise 
in the force tending towards absorption. In consequence of this 
sequence of events, considerable alterations of capillar}- pressure may 
be produced in impermeable capillaries, such as those in the limbs, 
without causing any appreciable increase in the lymph overflow from the 
limbs. On the other hand, where the capillaries are very permeable, 
very little pressure will be required to cause a transudation, since no 
work is done in the concentration of a proteid solution, and we find as a 
matter of fact, that capillaries where the pressure is lowest — i.e. in 
the liver — are also those which are the most permeable. Here, too, 
the absorbing force will be insignificant, since there is very little 
difference in the percentage of albumin between liver blood and liver 
lymph. 
Moreover, since the pressure on the venous side of the capillaries is 
considerably less than that on the arterial side, there will be a continual 
exudation of a very dilute lymph from the arterial capillaries, and a 
re-absorption of water and salts from this lymph in the venous 
capillaries. The lymph, therefore, will assume a composition such that 
the osmotic pressure of its proteids approximates the mean capillary 
pressure in the part where it is formed. 
This osmotic difference between blood plasma and tissue thud will 
not serve to explain the absorption of proteids by the blood vessels nor the 
absorption of serum from the serous cavities. It is difficult, however, if 
not impossible, to prove that serum or proteid is absorbed by the blood 
vessels. In some of iny transfusion experiments I have rendered a limb 
©edematous by means of serum, and in these cases have obtained no 
evidence at all of absorption by the blood vessels. There is no doubt 
that serum may be absorbed from the pleural and peritoneal cavities, 
but the absorption of these fluids is very much slower than the absorp- 
tion of salt solutions, and is, in fact, so slow that the whole of it can in 
most cases be effected by the lymphatic channels. A slow r absorption of 
serum from tissue spaces by means of the blood vessels is also physically 
possible. As the cells of the tissues feed on the proteids of the fluid, 
1 For a fuller discussion of this point, cf. Science Progress, London, 1896, vol. v. 
p. 151. 
