PR^ECIPITINS 471 



hands contained much less active opsonin towards the tubercle 

 bacillus than did the blood serum of a healthy individual. On 

 the other hand, the leucocytes of the tubercular girl could take 

 up as many tubercle bacilli as those of healthy serum when 

 normal blood serum was employed to furnish the opsonin. 1 



In conclusion, it may be stated that although the side-chain 

 theory is the only one which can be at present employed to 

 explain all the intricate reactions and interactions of immune 

 bodies, there possibly exists a much simpler relationship between 

 them. Indeed Arrhenius ( 8 ) and Madsen have recently shown that 

 many of the processes behave according to simple chemical laws 

 of neutralisation and solution. For example, the neutralisation 

 of tetanus toxin by its antitoxin behaves according to the same 

 laws as those which govern the neutralisation of a weak base by 

 a weak acid ; 2 and the action of varying amounts of agglutinin 

 on a suitable bacterium is analogous with the behaviour of a 

 soluble substance in varying amounts of two different solvents. 



The extreme importance of Ehrlich's theory as a means 

 of predicting what interactions of toxins and allied bodies 

 are possible cannot, however, be overestimated, and a thorough 

 mastery of the theory in all its details will well repay all who 

 possess it. It is the scaffolding from which the worker is able 

 to build up and piece together the disconnected material presented 

 to him. The complicated scaffolding must stand till the simpler, 

 but more durable, superstructure is completed. Without it, no 

 progress in the work is possible. 



The opsonic value of the serum is increased specifically by the 

 injection of minute doses of dead bacteria, e.g. T J^ grm. tuberculin. 

 (A negative phase precedes the positive phase.) Staphylococcic in- 

 fections (boils and acne) and tubercular lesions are being success- 

 fully treated by this method of vaccination. Auto-inoculation occurs 

 on infection, and this may be too much, leading to a persistent nega- 

 tive phase and death, or too little, leading to a chronic lesion. In 

 the first case auto-inoculation is limited by rest, i.e. by keeping the 

 circulation quiet in the infected part. In the second case fomenta- 



1 Barratt ( 20 ) has demonstrated that red corpuscles may be eaten by pha- 

 gocytes under the influence of an opsonin after removal of amboceptor from the 

 serum. The serum of a goat injected with sheep's red corpuscles caused pha- 

 gocytosis of rabbit's, sheep's and dove's red corpuscles. This proves that 

 opsonin and amboceptor are different substances. (Editor's Note.) 



2 This conclusion is controverted by experiments of J. Craw ( 21 ), who finds 

 he can separate a neutral mixture of a toxin and an antitoxin by means of a 

 gelatine filter, a result which is also not in agreement with Ehrlich's theory. 

 Craw thinks the union of antitoxin and toxin and bacteria with agglutinin 

 resembles that of a tissue with a dye. (Editor's Note.) 



