The importance of considering the effect of a possible flow out of the anterior chamber before inferring any mechanism of aqueous humor formation from the relative concentration of a substance in the aqueous humor and plasma under equilibrium conditions has been stressed. Several processes to account for the chemical equilibria between aqueous humor and blood based on the ultrafiltration and secretion hypotheses with a possible simultaneous loss of aqueous humor by flow have been outlined. On the basis of these processes, equations were formulated which would relate the rates of transfer into and out of the anterior chamber to the ratio of concentration of a substance in the aqueous to that in the blood at various intervals after its introduction into the blood. The explanation of equilibrium ratios above and below one for aqueous constituents is made apparent from the mathematical formulations. For each substance tested a determination was made of the best fit when the concentration in the aqueous humor is plotted against time. This fit was obtained by plotting the rate of transfer in against the rate of transfer out of the anterior chamber for all of the experimentally found concentration ratios on the basis of both the ultrafiltration and secretory hypotheses. Two sets of values were obtained from these calculations, one set for each hypothesis. The substantial agreement of all the experimental data with an assumed rate of leakage out of the anterior chamber of approximately 4 c. mm. per minute was shown to be compatible only with the idea that all the monovalent electrolytes tested entered the anterior chamber as a result of secretory process. It could not be decided from these chemical studies whether the non-electrolytes and the one multivalent electrolyte tested enter the anterior chamber by ultrafiltration or secretion. Experimental findings from other sources were cited which would suggest that non-electrolytes enter the anterior chamber as a result of ultrafiltration. The implications of the mechanism outlined in the paper with respect to intraocular pressure have been discussed. Supplementary evidence from the literature has been given in support of the conclusions presented here.