6 



SCIENCE. 



[N. S. Vol. XVIII. Xo. 444. 



Study of the iutravascular clotting of blood 

 caused by venom injections. We were 

 fortunate in having several kinds of venom 

 with which to undertake this investigation. 

 It has long been recognized that the venom 

 of the viper family, to which our rattle- 

 snake belongs, is especially prone to cause 

 vascular clotting. Besides crotalus venom 

 we have possessed other viper venoms- 

 such as that of Russell's viper from India 

 {Daboia Bussellii) and Trimere surus from 

 Japan. By far the most active venom in 

 this respect is that of Russell's viper, as 

 intravenous injections of it cause almost 

 instantaneous 'clotting' of the blood in the 

 right heart, pulmonary artery and vena 

 cavffi. If these clots are removed imme- 

 diately and examined (1) in the fresh 

 state, and (2) after instantaneous harden- 

 ing, no fibrin can be found. They consist 

 exclusively of masses of agglutinated red 

 corpuscles and are entirely free from evi- 

 dences of clotting in the classical sense. 

 But besides the change in adhesion of the 

 corpuscles, still greater alterations of form 

 and refraction have taken place in them; 

 and sections of the 'clot' from the heart 

 and in the vessels of the lung show ex- 

 traordinary pictures of drawn and twisted 

 bands of hyaline appearance which may 

 readily be mistaken for modified fibrin. 



These facts would seem to be of immense 

 assistance in explaining the origin of cer- 

 tain thrombi met with in man in infections 

 and other diseases, which are attended by 

 marked blood destruction, and perhaps still 

 other changes in the composition of the 

 blood. 



The light which the study of aggiutin- 

 ines has shed upon the general subject of 

 thrombosis has served also to illuminate in 

 no trifling way the path along which we 

 are proceeding in gaining knowledge of 

 certain forms of hfemorrhage. It is custom- 

 ary to ascribe liEemorrhages to two opposed 



conditions— to rupture of blood vessels and 

 escape of blood eorj^uscles through the ves- 

 sels by diapedesis. Only the second mode 

 of origin needs enlightenment. Hitherto 

 we have been obliged to be satisfied with a 

 vague and hypothetical molecular altera- 

 tion of the vascular wall as explaining the 

 increased passage of red corpuscles into the 

 tissues. It would now seem that at least 

 certain forms of parenchymatous haemor- 

 rhage are explicable in a more satisfactory 

 and objective manner; and I will ask you 

 kindly to turn your attention to this sub- 

 ject. 



That no necessary relation exists between 

 thrombosis and htemorrhage can be shown 

 experimentally by removing from rattle- 

 snake venom its agglutinines for red cor- 

 puscles, when the haemorrhage principle is 

 left unaffected; and that no relation exists 

 between haemolysis and hemorrhage can be 

 inferred from the action of cobra venom 

 which, while an active agent of blood de- 

 struction, causes but little hosmorrhage. 

 But the entire independence of the prin- 

 ciples which act injuriously upon blood 

 vessels, and thus permit escape of blood, 

 can also be shown by the use of riein, 

 which is non-hffimolytic, and which, in rab- 

 bits, produces extensive extravasation of 

 corpuscles into the serous membranes. 



Jacoby and Miiller have shown that riein 

 is not robbed of its entire toxicity by di- 

 gestion with artificial gastric juice, and 

 Miiller made the observation that the ag- 

 giutinine is destroyed m the process. I 

 found no trouble in confirming this finding 

 of Miiller, and in detecting that digested 

 riein is still capable of prodiicing hemor- 

 rhage. 



If the thin mesentery of the rabbit is 

 spread out and prepared and the capil- 

 laries lying within the hffimorrhagic areas 

 are carefully studied, it will be seen that 

 in neither venom nor riein poisoning have 



