1. Crystallized soy bean trypsin inhibitor, at a concentration of 100 µg./ml., suppressed the production of thrombin from a mixture of prothrombin and blood thrombokinase. The experiment was performed in the presence of 0.011 M oxalate, in order to minimize the possibility of participation by accessory factors which require ionic calcium. The results are in accord with the view that thrombokinase is a trypsin-like enzyme. 2. When a solution of blood thrombokinase was centrifuged at 85,000 g for 120 minutes, almost all the activity remained in the supernate. This supernate activated the supernate from a prothrombin solution which had been similarly centrifuged. The activation of prothrombin by thrombokinase can proceed in the absence of material completely sedimentable in 120 minutes at 85,000 g. 3. An "accelerator" reagent was prepared by treating bovine serum with barium carbonate, and then passing the serum through a column of diatomaceous earth. This "accelerator" was used together with prothrombin, blood thrombokinase, Howell's cephalin, and calcium chloride to compose a five-reagent thrombin-producing system. In this system, no thrombin was produced without thrombokinase. On the other hand, thrombin was produced from prothrombin and thrombokinase, even when all the other reagents were omitted. When calcium was omitted, thrombokinase was able to function; but cephalin and the "accelerator" reagent were ineffective. 4. Quantitative tests indicated that the "accelerator" reagent exerted an effect distinct from those of thrombokinase and cephalin. However, it is not certain whether the "accelerator" reagent functioned as an accessory factor, as a potential source of more thrombokinase, or both. In the experiments reported, thrombokinase was primary to, or necessary for, the effect of "accelerator." 5. The effectiveness of thrombokinase was multiplied a hundred times or more, when complemented by calcium, cephalin, and "accelerator" reagent. Ionic calcium was a necessary component of this complementing system. This may help to explain why removal of calcium ions keeps blood fluid, even though thrombokinase, by itself, is little influenced either by calcium ions or by oxalate.