1906.] Specificity and Action in Vitro of Gastrotoxin. 429 



the faintest trace of htemoglobin was seen to be diffused in the fluid 

 immediately above the deposited corpuscles. After the second injection, 

 however, it is found that guinea-pig's serum will reactivate the heated 

 immune serum to a considerable extent. In the above experiment six days 

 after the second injection 0"5 c.c. heated immune serum, on being reactivated 

 by 0'25 c.c. guinea-pig's serum, completely dissolved the test amount of 

 corpuscles. I have confirmed this experiment several times, and to my 

 mind it conclusively points to the presence of two distinct immune bodies. 

 (1) An increase of the normal hemolysin of the rabbit which is not comple- 

 mented by guinea-pig's serum. (2) A. newly-formed and therefore artificial 

 hsemolysin which is complemented by guinea-pig's serum. 



This has an important bearing upon the hypothesis of the multiplicity of 

 immune bodies, which is upheld by Ehrlich and Morgenroth (3), but is 

 denied by other observers, notably Muir and Browning (4), and Gay (5) 

 working in the Pasteur Institute. Agglutination of the red corpuscles also 

 occurs. 



2. Action upon the Protoplasmic Granules of the Gastric Cells. 



Method. — The method which I employ in order to demonstrate this action 

 is an imitation of that described for hemolysin. 



The mucous membrane of a guinea-pig's stomach is first washed free from 

 blood by sterilised salt solution, which is made to flow through a canula 

 introduced into the thoracic aorta, the stream issuing from the inferior vena 

 cava. It is then scraped off, and the pulp ground up in a glass mortar. An 

 emulsion is made with salt solution and centrifugalised for five minutes at 

 a low speed. The supernatant fluid on being pipetted off is found to contain 

 in suspension innumerable large and small protoplasmic granules. The 

 granules are separated from the albuminous fluid in which they float by 

 centrifugalisation at a high speed, and repeatedly washing until the 

 washings give no precipitate with potassium ferrocyanide and acetic acid. 



The final suspension of granules in saline solution, which is to be used, 

 must be well agitated so as to free all the granules, and before use it must 

 be slowly centrifugalised to idtimately remove any masses of granules or 

 pieces of tissue which happen to be present. 



A series of test-tubes is prepared, each tube containing 2 c.c. of the gastro- 

 toxic serum in increasing dilutions. The first tube contains undiluted 

 serum, and the remainder dilutions from 1 in 5 to 1 in 320. To the contents 

 of each tube three to five drops, according to concentration of the suspension 

 of granules in salt solution, are added. 



A control of normal saline is prepared, and also a control in which 



