540 BIOLOGY OF PNEUMOCOCCUS 



large quantity of physiological salt solution. In order to combat 

 these immediate reactions it is advisable to have on hand a supply 

 of epinephrine which may be injected into horses showing symp- 

 toms of possible hypersensitivity ; glucose solution may also be ad- 

 ministered intravenously to advantage. 



When living cultures are used as the immunizing agent, certain 

 complications may occur. White* observed four instances of pneu- 

 monitis, one of endocarditis, and one of spontaneous pneumo- 

 thorax in a group of horses in which living cultures had been 

 eventually substituted after a course of immunization with heat- 

 killed organisms. The pneumococci obtained from the lesions were 

 apparently identical with the strains routinely injected. An inter- 

 esting observation was that the serum of the horses at the time the 

 infections occurred contained sufficient protective antibody so that 

 0.2 cubic centimeter of serum prevented the death of mice injected 

 with 0.1 cubic centimeter of virulent culture, either of the organ- 

 isms isolated from the horse or of those used in preparing the vac- 

 cine. These apparently paradoxical phenomena may be explained 

 in part by the investigations of Griffith, 559 who reported that in 

 rabbits the effect of intravenous inoculations of culture into an 

 immune animal was to abolish temporarily the protective power of 

 the serum. Griffith thought that the transient suppression of anti- 

 bodies enabled the pneumococci to become established in the tis- 

 sues of the animals. It seems, however, that any alleged superiority 

 of serums from horses immunized with living cultures, unless the 

 dosage is very carefully controlled, would be outweighed by the 

 loss of animals from infections of the nature just described. 



BLEEDINGS 



Period of immunization prior to bleeding. The period of im- 

 munization necessary for the production in horses of sufficient pro- 

 tective antibodies to warrant initiating routine bleedings depends 

 on the material injected, the spacing of doses, and the response of 



* Unpublished observations. 



