Here another question presents itself,—do the two fluids accomplish their admixture in the capillary canals themselves, or do they cross the partition by different capillary canals, so that neither fluid mixes with its opposite fluid until the moment of its exit from the capillary canals ? On the latter hypothesis it were necessary to admit that the number and diameter of the capillary canals followed sepa rately by each of the two fluids must be per fectly equal, for, without that, how would the general result of this double permeation, a result which is explained by the quantity of endosmo sis, be in exact relation with the capillary action on the two fluids ? Now it is repugnant to reason to admit any such perfect equality among all the capillary canals, or to suppose an equal number especially fitted for the transmission of each of the two fluids. It must then necessarily be allowed that the transmission of the two op posite fluids takes place by the same capillary canals, and that consequently this double movement of transmission takes place by a reciprocal penetration of the two fluids.
The preceding theory, with which I was at one time inclined to rest satisfied, and which, indeed, seemed to be based on a sufficiently firm foundation, was however brought into jeo pardy by a discovery which I made subse quently, in regard to the phenomena of endos mosis exhibited by certain acids separated from pure water by a layer of animal mem brane.
In the earliest experiments 1 made on the endosmosis of the acids, I observed a number of anomalous phenomena, for which I felt my self incompetent to assign any sufficient reason. I had always placed the acids above the water, from which they were separated by a layer of animal membrane. Certain acids, such as the hydrochloric, at very different degrees of den sity, and nitric acid only at pretty high degrees of density, gave me an endosmosis, the current of which was directed from the inferior water towards the superior acid, so that the acid rose gradually in the tube of the endosmometer. On the other hand, I had always found the sulphuric acid pretty largely diluted, and the hydrosulphurie acid, under the same circum stances as the acids mentioned above, gradually to sink in the tube of the endosmometer. con cluded from this that these acids did not occasion any endosmosis, and that they passed mechani cally, and merely in virtue of their gravity, athwart the animal membrane to mingle with the water. I had also found that the sulphuric and hydrosulphurIc acids, added to gum-water, deprived it of the faculty ofproducing endos mosis, and that this acidulated water fell in the tube of the endosmometer, instead of rising, as simple gum-water constantly does. These facts induced me to say metaphorically that the sul phuric and hydrosuiphuric acids were the ene mies of endosmosis.
More recent inquiries have enabled me to see the above mentioned phenomena in ano ther light. It was the oxalic acid led me to the conclusions I shall now lay be fore the reader. Having poured a solution of this acid into the endosinometer closed with a piece of bladder, and placed the re servoir in water, I found the acid fluid sink rapidly in the tube, and flow towards the inferior water, making its way by filtration through the animal membrane. 1 then reversed the arrangement observed in this experi ment. I tilled the endosmometer with water, and plunged the reservoir into a solution of oxalic acid. I was now surprised to find the
water making its way rapidly into the endos mumeter, and the column rising in the tube, so that, in opposition to all I had yet observed, here was the current of endosmosis directed from the acid towards the water. The follow ing are the particulars of this experiment. I laving poured some rain-water into the reser voir of the endosmometer, I plunged the reser voir, closed with a piece of bladder, into a so lution of oxalic acid of the density of 1.045, (11.6 parts of crystallized acid in 100 of the solution,) the temperature being -I- 25 cent. The ascent of the water in the tube of the en &milometer lasted for three days, becoming gradually slower and slower. The ascent hav ing then become almost imperceptible, I emp tied the endosmometer, in the contents of which I found water charged with oxalic acid. The exterior fluid was reduced in density to 1.033, so that, whilst the lower acid had pene trated the upper water by endosmosis, the water had penetrated the acid by exosmosis, and thus diminished its density ; the permea tion of the water, however, had been less con siderable than that of the acid ; so that the upper water, increased in volume, had risen in the tube of the endosmometer. We have thus, in the present instance, another obvious proof of the existence of two opposite and unequal currents. Having filled the reservoir of the en dosmometer anew with rain-water, I placed it in the solution of oxalic acid already used, and of the reduced density of 1.033. The ascent in the tube which again occurred, having almost ceased at the end of two days, I tested the fluid in the endosmometer, and found it to con tain oxalic acid,and discovered the density of the external fluid further reduced to 1.025. I re peated the same experiment a third time, filling the reservoir of the endosmotneter with rain water, and plunging it in the old acid solution. Endosmosis went on as before, but with less celerity. having given up the experiment, after the lapse of twenty-four hours I found the density of the exterior fluid now reduced to 1.023, and the internal fluid to contain a por tion of oxalic acid as before. I reduced the density of the exterior acid solution to 1.01, but the included water still gave evidence of a pretty active endosmosis. 1 reduced the den sity of the acid to 1.005, (1.2 of acid to 100 of the solution,) and the endosmosis was still very remarkable. In these experiments I found that the endosmosis was by so much the more rapid as the exterior acid solution was more dense, so that the capacity of oxalic acid to permeate an animal membrane would appear to increase with the density of its solution in water. In these experiments, too, we observe a fluid, more dense than water, and having a less power of capillary ascent than it, never theless forming the stronger current, or current of endosmosis, whilst the water opposed to this fluid forms the weaker current, or counter-cur rent of exosmosis. This is in opposition to all I had observed before ; and the theory I had raised on the different capacities of capillary ascent possessed by two opposed fluids is con sequently shaken, or at all events proved to be no longer generally applicable. What may be the cause of this new phenomenon? Do animal membranes give passage more readily through their meshes to solutions of oxalic acid than to water ? This point I sought to determine by the following experiments.