In order to given an experimental confirmation to the conclusions arrived at in the firsttwo notes of a series of the present study, the author has, like Rossle-Eppinger, made an injury into the serum protein picture in hepatic disfunction brought about by allyl formate injection into dogs. We have also estimated the serum protein in the similarly treated dogs whose blood vascular wall permeability has been subjected to influence by administration of adrenocortical hormones, ACTH, etc. This latter experiment was conducted with an objective of obtaining a suggestion for therapy. When the hepatic function is rapidly impaired by an injection of 500 mg of allyl formate for each 10 kg of body weight, the serum albumin becomes decreased γ-globulin increased, and the fluctuation in these changes in terms of time elapsed showed an inverse relationship. When the hepatic impairment is brought about gradually by daily injections of 250 mg of allyl formats for each 10 kg of body weight, the albumin was at first reduced, γ-globulin markedly increased, α-globulin also showed an increase during the period in whichthe albumin is greatly increased but returns to the normal level at later stages. This result may be regarded as a rise in permeability of the blood vascular wall, and supports the theory of serous inflammation as proposed by Rossle-Eppinger. An injection of adrenocortical hormones like Cortison and DOCA, ACTH and glucuronic acid preparations to the dog whose hepatic disfunction has been brought about by treatment inter allyl formate caused a certain restriction on the decrease of albumin and increase of γ-globulin as compared to the uninjected animal. This fact suggests that these drugs inhibit to some extent the increase in permeability of the blood vascular wall otherwise, and for this reason may be considered to pass a beneficial action in corresponding diseases. It was further made clear that the intensity of such an effect is roughly iu the order named ACTH, Cortison, DOCA, glucuronic acid preparation.