Contagious Pleuro-Pneuononia of Cattle. 181 



Complement fixation tests were carried out, and a description, 

 was given of a complement fixation test which has been applied 

 for the diagnosis of contagious pleuro-pneumonia in cattle. The 

 technique of this test, however, is considered to be too intricate 

 and laborious to allow of its adoption as a routine diagnostic 

 method. Further, it was found that the test was only appro- 

 ximately accurate in its results, for, on testing the sera of 63 

 different animals (cattle) a positive result was obtained with the 

 serum of one animal which, on subsequent post-mortem examina- 

 tion, showed no lesions of contagious pleuro-pneumonia in the 

 lungs. Two other animals gave reactions which could not be 

 definitely interpreted either as negative or positive. On sub- 

 mitting these two animals to post mortem examination, no lesions 

 of contagious pleuro-pneumonia were discovered. 



The results of the tests of these 63 serum samples can be- 

 conveniently tabulated as follow : — 



Number of serum samples tested — 63 



Gave positive reaction to test and showed lesions of C.P.P. 



on P.M. ...... 13 



Gave positive reaction to test and showed no lesions of 



C.P.P. on P.M. ..... 1 



Gave neg^ative reaction to test and showed lesions of 



C.P.P. on P.M. 



Gave negative reaction to test and showed no lesions of 



C.P.P. on P.M. 47 



Gave border line reaction to test and showed lesions of 



C.P.P. on P.M. ..... 



Gave border line reaction to test and showed no lesions of 



C.P.P. on P.M. ..... 2 



At first sight this tabulation appears to show that the test: 

 has been fairly accurate in differentiating between animals which 

 were and which were not affected with contagious pleuro-pneu- 

 monia, but if the figures are analysed carefully, it is found that 

 the percentage of error is an unduly large one. Sixty-three- 

 serum samples were tested, and of the reactions obtained, 60 

 were verified by the post-mortem findings, while the other three 

 were not. In a total of 50 negative sera tested, 47 reacted nega- 

 tively, and three gave reactions which were not negative, an error 

 of 6 per cent. Fourteen serum samples gave positive reactions; 

 13 of these were verified by the post-mortem findings, one was 

 not; an error of approximately 7.14 per cent. If we add to these 



