This section is from the book "Animal Physiology: The Structure And Functions Of The Human Body", by John Cleland. Also available from Amazon: Animal Physiology, the Structure and Functions of the Human Body.
The Passage Of Liquids Through Membranes is regulated by physical laws of diffusion, which are closely connected with capillary attraction. Just as gases diffuse according to definite laws, so also do liquids. Their diffusion through membranes or porous septa is called osmosis; or, inasmuch as there are two currents in opposite directions wherever a membrane separates two different fluids, the words endos-mosis and exosmosis may be used to indicate the inward and outward flow. If a piece of moist bladder be stretched across a tube, and any saline solution introduced into the vessel thus made, and the end of the tube be then dipped in water, it will be found that in a short time a portion of the solution has passed through into the water, while a larger amount of water has passed into the tube, and raised the height of the liquid within it. The same experiment may be made with solutions of different sorts on the two sides of the membrane. But the important points to note are, that different solutions pass through in definite proportion to the amount of any particular substance passing in the opposite direction, and that while some substances diffuse with facility, others do so with difficulty. The substances which diffuse easily are called crystalloids, while those which diffuse with difficulty are called colloids (Graham). Thus albumen in its ordinary condition is a colloid, but when converted into peptone it becomes crystalloid. It is in consequence of endos-mosis that, when water is added to blood, the red corpuscles become swollen and spherical. The substance in which the fluid parts of the corpuscle are entangled acts as a membrane would, and while a certain amount of fluid passes out, a larger amount of water passes in and gorges the corpuscle. So also nucleated cells, when water is added to them, become rapidly swollen, till they burst and are destroyed.
Now, there seems no reason to doubt that the absorption into the capillary blood-vessels is an instance of endosmosis without intervention of any vital force. All sorts of salts and other diffusible substances, whether simply useless or positively injurious, find their way into them; while it is proved by experiment that such substances do not pass into the lacteals, at least so rapidly. And this is not altogether inexplicable; for we have seen that the capillaries are near the surface of the villus, while the lacteal is in the centre, and receives its supplies through the action of nucleated corpuscles. In fact, a little reflection will show that the action of the epithelium in lacteal absorption differs from that of secreting cells in separating substances from the blood in nothing, save that in secretion the current is from vessels to a free surface, while in lacteal absorption the current is from a free surface to the interior of a vessel.