168 APPLIED BACTERIOLOGY 



January, 1896) has arrived at the conclusion that the blood 

 serum of normal horses may possess antitoxic power, and 

 that there are two distinct therapeutic agents present in the 

 blood of immunised horses. He also finds that there is a 

 gradual diminution of antitoxic power in the serum 

 yielded by horses, even though they continue to receive 

 doses of toxins. He is of opinion that the best method of 

 obtaining antitoxic serum is to begin with injecting the 

 horses with a culture of living bacilli. 



The following is an extract from an interesting and 

 important report by Dr. J. W. Washbourn upon the serum 

 treatment of diphtheria, issued to the Medical Congress 

 held in London during August, 1895 : 



' The value of the antitoxin treatment has been tested in 

 two ways : the one is the clinical method, the observation 

 of the effect of antitoxin upon the course of the disease ; 

 and the other is the statistical method, the comparison 

 of the mortality of a series of cases treated by antitoxin 

 with that of similar series treated in other ways. 



'In order to arrive at a correct conclusion by either 

 method, certain precautions must be taken, and it will be 

 my duty to point out the fallacies that may arise. 



' Let us first consider the statistical method ; and here I 

 may say that I only propose to treat the matter from a 

 general point of view, for I have been given to understand 

 that the details of the statistics are to be discussed in 

 another section. We know that all statistics are open to 

 fallacies if the number of cases is not sufficiently large, and 

 if other precautions are not taken into account. 



' In diphtheria the age of the patient is a most important 

 factor in determining the mortality. The statistics of the 

 Metropolitan Asylums Board clearly demonstrate this point. 

 Out of 4,435 patients under 5 years of age admitted into 

 the Board hospitals between the years 1888 and 1894, the 



