564 BIOLOGY OF PNEUMOCOCCUS 



Tests for chill- producing properties. Until the nature of the 

 chill-producing factor is known and until the substance can be re- 

 moved from antipneumococcic serum, it is necessary, prior to dis- 

 tribution of the product, to have recourse to some means of deter- 

 mining the presence or absence of this property. It is not expedi- 

 ent to make preliminary tests on human subjects, and the smaller, 

 more commonly used laboratory animals are unsatisfactory for 

 the purpose. Sabin and Wallace 1209 stated that it was possible to 

 differentiate chill-producing from satisfactory lots of serum by in- 

 travenous injection into suitable dogs. A rise in temperature of 

 1.5°F. or more was considered the criterion of a positive reaction, 

 regardless of the presence or absence of a definite chill. In a later 

 and more complete paper, Sabin and Wallace 1210 reported observa- 

 tions on a considerable number of serums tested by the method, and 

 confirmed their previous results. Barnes and Kramer (1933) 83 

 made a preliminary report on the use of monkeys (Macacus rhe- 

 sus) in detecting chill-producing lots of concentrated serum and 

 found the response of the animals to correspond closely with those 

 observed in patients with lobar pneumonia treated with the prepa- 

 rations. In the test animal, the presence or absence of frank chills 

 appeared to be of more significance than elevation of body tem- 

 perature. 



In comparing the products obtained by the sodium sulfate and 

 alcohol methods, Barnes and Kramer 83 tested a pooled lot of anti- 

 pneumococcic serum that had been divided into two equal parts. 

 One batch had been concentrated by sodium sulfate precipitation, 

 the other by alcohol. The sodium sulfate concentrate gave chills in 

 a high percentage of cases, while the batch refined with alcohol was 

 essentially chill-free in its action. 



The practical value of the method was affirmed by the observa- 

 tions of Barnes and Robinson (1936), 84 who found that applica- 

 tion of the test over an extended period of time and the rejection 

 of lots of serum eliciting chills in the test monkeys had reduced the 



