The Action of" Peptone" on Blood andsjmmunity thereto. 371 



it. So far as the present writers are aware this criticism applies to all 

 experiments in which " peptone " has been injected into animals with damaged 

 or more or less excised livers. 



Of the facts unexplained by the hypothesis that the secretion by the' liver 

 of antithrombin is the cause of the fluidity of normal circulating blood and of 

 the retarded coagulability of peptone blood, it is noteworthy that the massive 

 injection of thrombin does not produce thrombosis, Wooldridge (13). This was 

 confirmed by Mellanby (14) who injected, without thrombosis, sufficient 

 thrombin into a cat to coagulate 2 litres of blood. After the intravascular 

 injection of thrombin, Davis (15) found the coagulation time of shed blood 

 to be variable, sometimes slightly lengthened at other times shortened. 

 Unfortunately Davis recorded as " coagulation times " only the time of 

 completion of the clot. This may be a variable period when the times of 

 commencement of clotting are approximately constant (Nos. 1-4). Moreover, 

 Davis' variations in coagulation times are not greater than can be accounted 

 for by alterations in the relative concentrations of carbon dioxide and oxygen 

 in the blood shed. Apparently this and the corresponding hydrion concen- 

 tration did not engage his attention. Davis explains the slightly decreased 

 coagulability following the injection of thrombin by suggesting that thrombin 

 produces an excessive secretion of antithrombin, while Howell (16) maintains 

 that thrombin, or prothrombin is a hormone which stimulates the secretion of 

 antithrombin. Davis frankly admits that if antithrombin exists in the 

 circulation and is a genuine antibody, secreted as a protection against thrombin, 

 then according to all analogies the presence of thrombin should lead to its 

 production. 



Nolf (17), however, denies that the intravascular injection of thrombin leads 

 to the formation of antithrombin. Further doubt is cast on the accepted 

 hypothesis by the recent work of Arthus (18), who observed that delay occurs 

 in the coagulation of shed blood after the injection of the venom of Crotalus 

 adamanteus into immunised, non-immunised, and even anaphylactic rabbits. 

 This not only happened in whole animals, but in those in which the hepatic 

 blood supply had been cut off. 



Two views have been advanced in this connection, viz., that of Martin (19), 

 and of Barratt (20), where the active fraction of snake venom is considered to 

 be a true fibrin-ferment (thrombin), and that of Mellanby (21), that it is a 

 kinase which, when rapidly injected, causes the rapid formation of thrombin. 

 It is evident from the experiments of Arthus, that the " negative phase " of 

 coagulation following either the rapid injection of minimal doses of venom, or 

 the slow injection of larger doses, is not produced by the secretion by the 

 liver of antithrombin, which combines with thrombin, either existing in, or 



2 D 2 



