METHODS OF EXAMINATION. 221 



characteristic of a non-fatal case of the disease takes place when 

 the balance of antitoxin against toxin is in favour of the former. 

 The pneumococci after the crisis, as has been proved both 

 -culturally and by inoculation experiments, are still vital and 

 "virulent, though not so virulent as when the fever is at its height. 

 On them directly the antitoxin has no effect, but any toxin now 

 elaborated by them is neutralised, and has no longer either local 

 or general pathogenic effects. 



A fact interesting as corroborating the view that the pneumo- 

 coccus is really the cause of acute lobar pneumonia is that the 

 serum of patients who have recovered from pneumonia has 

 in a certain proportion of cases a protective effect against the 

 pneumococcus in rabbits. So far as our knowledge goes, such 

 a protective serum is specific, or, in other words, protects only 

 against the organism by the action of which its protective 

 properties have been produced, and therefore it must be against 

 the pneumococcus that the human subject requires protection 

 in pneumonia. 



The Klemperers treated a certain number of cases of human 

 pneumonia by serum derived from immune animals, and appar- 

 ently with a certain measure of success, and Washbourn's serum 

 has also been used. Although the use of these sera apparently 

 causes the temperature to fall, and in some cases appears to 

 hasten a crisis, further experience is necessary before their value 

 in therapeutics can be properly estimated. 



If a small amount of a culture of Fraenkel's pneumococcus 

 be placed in an anti-pneumococcic serum, an aggregation of the 

 bacteria into clumps occurs. Such an agglutination, as it is 

 called, is frequently observed under similar circumstances with 

 other bacteria. The phenomenon is not invariably associated 

 with the presence of protective bodies in a serum, but it has 

 been used for diagnostic purposes in the differentiation of sore 

 throats due to pneumococcus infection from those due to other 

 bacteria. Whether the method is reliable has still to be proved. 



Methods of Examination. — These have been already de- 

 scribed, but may be summarised thus: (i) Microscopic. Stain 

 films from the densest part of the sputum or from the area of 

 spreading inflammation in the lung by Gram's method and by 

 carbol-fuchsin, etc. (p. 102), in the latter case without decoloris- 

 ing the groundwork of the preparation. 



