278 MANUAL OF MICROBIOLOGICAL METHODS 



Thus in the case of the influenza viruses, complement fixation readily 

 succeeds in placing a number of strains within a given type while separat- 

 ing one type from another quite sharply. The antihemagglutination 

 test on the other hand is less sensitive in recognizing widely separated 

 strains as belonging to the same type but quite readily shows differences 

 among strains. 



A rise in antibody titer during an infection may be elicited by the 

 complement-fixation technic using paired serum specimens much as in the 

 antihemagglutination test. Positive complement-fixation tests may be 

 obtained earlier or later than positive results using other serological 

 methods. The complement-fixing antibody may be detectable for longer 

 or shorter periods of time than antibody revealed by other methods. 

 These last two peculiarities vary with the disease in question. 



Neutralization 



General observations. In vitro neutraUzation tests, e.g., hemag- 

 glutinin inhibition, may or may not be directly related to in vivo neu- 

 tralization. In vivo neutralization tests vary with the virus and the 

 living cells to be protected. Not uncommonly among virus diseases 

 active immunity seems to transcend circulating antibody as an entity. 

 Nevertheless neutralization tests are useful in assaying the protective 

 power of circulating antibody. Generally speaking an in vivo neutraliza- 

 tion test has as its goal the determination of the amount of specific 

 antibody required to prevent the virus from producing observable 

 changes in host cells or to prevent viral propagation. 



Ingredients. It should be remembered that inactive virus can bind 

 antibody, but only active virus enters into estabhshing infection. There- 

 fore, it is easy to imagine disparities in experimental results (McKee 

 and Hale, 1947). A preparation containing nothing but active virus 

 would be desirable but probably impossible to achieve. Harvesting the 

 virus at an optimum time and using it immediately minimize the ratio 

 of inactive virus to active virus. If nonviral antigen adheres to the virus 

 or is present in the mixture, neutrahzation tests may be in error. For 

 example, if virus is produced in the egg and rabbit antiserum to egg virus 

 is used, at least two antigen-antibody systems are involved. Undoubtedly 

 this problem can be minimized by using virus and antibody from the 

 species used as the test animal. This is not always convenient or possible. 



Dilutions. Either the antibody or the virus may be diluted while the 

 other is held constant. As with other antigen-antibody systems the 

 straight-line function fails when too much or too little of either reagent is 

 employed. Biological variation is great enough that serial dilutions 

 by less than mcrements of 10 may produce equivocal end points. This 

 is not invariably true, for doubling dilutions sometimes produce accept- 



