856 



SCIENCE. 



[N. S. Vol. XVI. No. 413. 



the operation of substances agglutinative 

 of red corpuscles. Probably every one 

 with large experience in the examination 

 of fresh blood in disease has noticed that 

 sometimes red corpuscles, examined im- 

 mediately after withdrawal of the blood, 

 have a peculiar tendency to form clumps 

 which cannot readily be broken up. This 

 phenomenon, which is certainly suggestive 

 of the action of an agglutinating agent, I 

 have observed especially in some cases of 

 septic infections and of cirrhosis of the 

 liver. 



Furthermore, I would emphasize the sup- 

 port given by the recognition of liasmag- 

 glutinins to views advocated many years 

 ago by Hueter and by Klebs concerning the 

 occurrence of thrombi composed of coa^ 

 lesced red blood corpuscles. Such thrombi 

 I believe to be not uncommon in typhoid 

 fever and, other infections, especially in 

 small blood vessels. I have elsewhere called 

 attention to the evidence in favor of the 

 interpretation of many of the hyaline 

 thrombi as derived from agglutinated red 

 corpuscles. 



It can scarcely be doubted that sub- 

 stances agglutinative of white blood cor- 

 puscles are also produced by certain bac- 

 teria, and that these are concerned in the 

 clumping- of pus cells and of leucocytes 

 within the living body, but it would not be 

 profitable to discuss this matter without 

 more exact information than we now 

 possess. 



In this connection I may say that not 

 only the discovery of the bacterial htemag- 

 glutinins, but also that of the hfemolysins 

 and the leucolysins, is likely to shed new 

 light upon certain aspects of the difficult 

 subject of thrombosis. The red corpuscles 

 undergo various moi-phological changes un- 

 der the influence of different bacterial 

 lia?molysins acting with varying intensity. 

 Distortions of shape, throwing out of pro- 

 jections, and detachment of colorless par- 



ticles resembling platelets, can sometimes 

 be seen. These observations are of special 

 interest . with reference to the doctrine, 

 already strongly supported, that platelet 

 thrombi originate from disintegrated red 

 corpuscles. Levaditi, and Neisser and 

 Wechsberg, have described, as the result 

 of intravenous injections of Staphylococcus 

 aureus, areas of coagulative necrosis in the 

 rabbit's kidney, which they attribute to 

 thrombi composed of disintegrated leu- 

 cocytes caused by the staphylococcus leu- 

 cocidin, to which I have already referred. 

 I 'have dwelt at some length, although of 

 necessity incompletely, upon the bacterial 

 ha?molysins, leucocidins and hajmagglutin- 

 ins, because we are better' informed about 

 these agents than concerning other mem- 

 bers of this recently recognized class of 

 bacterial toxins. I have already expressed 

 the opinion that similar poisons acting spe- 

 cifically i;pon other cells of the body are 

 produced by bacteria ; indeed neurotoxins 

 and nephrotoxins of this type have been 

 reported. The difficulties in the way of 

 direct proof of the existence of these other 

 bacterial cytotoxins are greater than in the 

 case of those acting upon the red and the 

 white blood coi-puscles, but doubtless they 

 can be overcome. Of course we have evi- 

 dence of the action of bacterial poisons 

 upon various body cells, but this is not 

 enough. At present we can apply only in 

 a vague and unsatisfactory way to the ex- 

 planation of pathological processes most of 

 the knowledge of this kind which we pos- 

 sess. What is urgently needed is a separa- 

 tion of these poisons and a determination 

 of their source, constitution, mode of action 

 and degree of specificity along such lines 

 as have been followed so fruitfully in the 

 investigation of the soluble diphthei'ia and 

 tetanus toxins, those other toxins of bac- 

 teria and of venom already considered, 

 and the cytotoxins of normal and of im- 

 mune serum. The path leading apparently 



