Walker Hall, discussing this same question, holds that there is but little evidence “as to the relation of uricæmia to the formation of tophi.” He asks the question whether the deposition is the outcome of abnormal purin combination in the blood and lymph stream? which latter at present, he states, are regarded as passive carriers of the urates. For, he says, the small purin increase in gouty blood cannot surely make all the difference, seeing the large volume of solubility still available. The physico-chemical hypothesis, he claims, is inadequate to explain the relationship between uricæmia and the tophi, and hazards the suggestion that after all it may be that “the uricæmia plays little or no part in the depositions, and that these are due to the defective removal of substances resultant from local nuclear activities.” He asks, moreover, whether such substances differ in type from those of normal nuclein metabolism and so fail to be suspended in the surrounding lymph in such a way as to ensure their entrance into the blood-stream? Like others, he notes that atophan brings about a removal of some of the deposited urates. But such diminution of the tophi may, of course, he says, be due to increased flow of serum to the inflamed part; though, on the other hand, the more massive deposits “are surrounded by layers of young granulation tissue and phagocytes and peritophal fibrous tissue, and these in turn offer some hindrance to the permeation of serum or drugs.” In reviewing the foregoing views as to the formation of tophi, it is obvious we stand in urgent need of more knowledge. Neither the chemical nor the physical theory or a combination of the twain seems adequate. This for the salient reason that, as far as the existing evidence permits us to draw conclusions, it would seem probable that not only local but constitutional or systemic conditions play an important _rôle_ in tophi formation. But as far as our discussion has advanced, we may, we think, be justified in the following deductions:— (1) That tophi are blends of crystalloids and colloids and subject to the complex conditions of solubility attaching to such combinations. (2) That the relatively high sodium content of certain tissues, _e.g._, cartilage, favours the incidence of uratic deposits therein. (3) That tissue necrosis is not necessarily an antecedent to uratic deposits. (4) That no relationship can be established between the incidence or multiplicity of tophi and uricæmia. It will be seen from these conclusions that the proximate cause responsible for the genesis of tophi is yet to seek, and in pursuance of our quest we turn to another aspect of this complex subject. TISSUE AFFINITIES FOR URIC ACID