The following findings are incompatible with the theory that the intra-ocular fluid is a simple dialysate in equilibrium with the capillary plasma. The concns. of Cl and Na in the aqueous are too high. Nitrogenous substances, e.g., urea, creatinine, etc. penetrate the blood-aqueous barrier very slowly and have a low final concn. in the aqueous despite their relatively small molecules. Sucrose behaves similarly. The concn. of vit. C in the aqueous is too high but this may be accounted for by its synthesis in the lens. The aqueous osmotic pressure (0. P.) is greater than that of the plasma. There is an absence of an invariable and proportional variation in intra-ocular pressure in response to changes in blood pressure (B. P.) or colloid O. P. difference between plasma and aqueous. However, a vascular compensatory mechanism may exist whereby the capillaries do not necessarily follow changes in the general level of B. P. and there may also be a compensatory mechanism concerned with the drainage of intraocular fluid, hence this point may not be so significant. In other respects the aqueous has the characteristics of a dialysate. Two theories are put forward: (1) the aqueous may be primarily an ultra-filtrate of plasma with a superimposed activity of the ciliary epithelium which modifies the concns. of selected substances in line with the above findings; (2) the aqueous may be elaborated entirely within the ciliary epithelial cells with subsequent modification in its constituents when it circulates in the anterior chamber. The former theory appears the more likely.