PR^CIPITINS 467 



of horse flesh, but not in extracts of ox flesh. In this way horse 

 flesh can be distinguished from ox flesh. 



Although Ehrlich's theory shows us how anti-bodies in general 

 may be produced, it does not explain why certain bacteria, such 

 as those of typhoid fever and cholera, should exercise a toxic 

 action on the tissues, when neither of these bacteria liberates any 

 toxins into the culture fluids in which they are grown. The 

 side-chain theory tells us that anti-bodies against bacteria are 

 produced by a process analogous with that which produces hse- 

 molysins, but it leaves unexplained how these bacteria themselves 

 poison the tissues. Radzievsky ( 15 ) has tried to explain their 

 action by supposing that they are destroyed in the body, and that, 

 as they fall to pieces, the toxins which act on the tissue cells 

 are liberated. As Welch has shown, however, such a mechanism is 

 impossible, and to explain their toxic action this worker has 

 suggested another hypothesis. He points out that all invading 

 cells (bacteria, erythrocytes, epithelial cells, spermatozoa, &c.) 

 are , acted on by the tissue cells in the same way as these 

 latter act on the bacteria, &c. ; that is, by the intermediation 

 of receptors. The bacterium, he supposes, possesses a variety of 

 receptors, some of which are fitted by receptors of tissue cells, 

 and are, therefore, rendered functionless to the bacterium. As a 

 consequence of this, then, the bacterium reproduces its lost re- 

 ceptors in over excess, and sheds them into the plasma, where, 

 obeying the same laws which apply to tissue receptors, they will 

 anchor on to the tissue cells of the host, and, by bringing the 

 complement into contact with these, will thereby poison them. 



" It may perhaps aid in grasping the ideas here presented to 

 imagine the bacterium in the capacity of the host, as a structure 

 so large that one could inject animal cells into it. Provided the 

 proper receptor apparatus be present, the resulting reaction on 

 the part of the bacterium, as described, would be a process of 

 immunisation against the animal cells, through the intermediation 

 of specific cellulicidal substances" (Welch). 



This reciprocal reaction of tissue cell and invading cell might 

 be represented by a formula somewhat on the same plan as that 

 suggested by Sollmann ( 16 ). Let H represent the cell of the host, 

 F the invading cell, R the amboceptor attached to H, R that 

 attached to F, and C the complement. We have seen that the 

 complement C is present in normal serum, and that it is unaffected 



