Page:Elementary Text-book of Physics (Anthony, 1897).djvu/119

§ 94] placed in a vertical position, the liquid will rise in the tube until a very considerable pressure is attained. Dutrochet called such an instrument an endosmometer.

Graham made use of a similar instrument, which he called an osmometer, by means of which he studied, not only the action of poraus substances, such as are mentioned above, but also that of various organic tissues; and he was able to reach quantitative results of great value. Pfeffer has more recently made an extended study of the phenomena of osmosis, especially in those aspects relating to physiological phenomena. He has shown that colloid membranes produced by purely chemical means are even more efficient than the organic membranes employed by Graham.

93. Dialysis.—Upon the principles just set forth Graham has founded a method of separating crystalloids from any colloid matters in which they may be contained, which is often of great importance in chemical investigations. The apparatus employed by Graham consists of a hoop, over one side of which parchment paper is stretched so as to constitute a shallow basin. In this basin is placed the mixture under investigation, and the basin is then floated upon pure water contained in an outer vessel. If crystalloids be present, they will in due time pass through the membrane into the water, leaving the colloids behind. The process is often employed in toxicology for separating poisons from ingesta or other matters suspected of containing them. It is called dialysis, and the substances that pass through are said to dialyse.

94. Osmotic Pressure.— Pfeffer carried out a series of investigations to determine whether osmosis produces inequalities of pressure on the opposite sides of the membrane. In his examination of this question Pfeffer used membranes formed by chemical action in the pores of earthenware cells. These cells, which were designed to hold the solution to be examined, could be tightly closed and connected with a manometer, or instrument for measuring pressure. In the typical cases examined, the membranes permitted the solvent to pass through them freely, but were impervious to the