THE H^MORRHAGIC DIATHESIS 75 



prepared a solution of fibrinogen from normal or 

 haemophiliac plasma in the ordinary way by precipi- 

 tating it by passing a stream of carbon dioxide through 

 plasma kept from clotting by citrate or oxalate. 

 Fibrinogen so obtained, as Mellanby shows, always 

 carries with it prothrombin, and in the presence of 

 calcium salts and thrombokinase would liberate 

 thrombin. Addis, however, added instead a trace of 

 thrombin, which clotted the fibrinogen and left its 

 prothrombin in solution. When a trace of prothrombin 

 so obtained from a normal blood was added to haemo- 

 philiac blood, this promptly coagulated. The criticism 

 would of course be that there was some unused 

 thrombin present as well, too much having been added 

 to the fibrinogen. 



Thus, the exact pathology of haemophilia would be, 

 in Addis' s opinion, a congenital defect in the con- 

 stitution of the prothrombin, whereby it yields 

 thrombin much too slowly. Possibly the leucocytes 

 are ultimately at fault. 



The practical deduction we shall see later. 



5. There is no excess of antithrombin in the plasma 

 of the bleeder. If there were, the addition of a trace 

 of normal blood would not cause haemophiliac blood 

 to clot as it does, because any thrombin in the former 

 would be overpowered and destroyed by the anti- 

 thrombin in the latter. 



To sum up, the secret of haemophilia lies in a 

 defective quality of the prothrombin, such that it 

 takes much longer than usual to develop into 

 thrombin. No evidence is yet to hand to show 

 whether the haemorrhagic tendencies in scurvy, 



