166 THE AMERICAN MONTHLY [May 



preventing- outbreaks of epidemics of diphtheria, which 

 in past winters had occurred in this hospital with most 

 disastrous results. 



Out of 316 cases examined, 26 at one time or another 

 showed the presence of the Klebs-Loeffler bacillus. Two 

 of these had clinical diphtheria, so that out of 314 normal 

 throats and noses, 7.9 per cent contained the bacillus of 

 diphtheria. The average persistence of the bacillus on 

 the mucous membrane was fifteen days; the shortest 

 period one day, the longest 103 days. The nose was the 

 principal habitat in 65 per cent and the throat in 35 per 

 cent. The degree of virulence possessed by the bacilli in 

 the various cases was not determined. 



Antitoxin in Diphtheria. — The American Pediatric 

 Society are about to undertake a second collective investi- 

 gation of antitoxin, and they now ask that records of cases 

 of diphtheria involving the larynx, whether operated or 

 not, occurring in the United States, be sent to the Secre- 

 tary, W. P. Northrup, M.D., 57 East Seventy-ninth street. 

 New York, N. Y. 



The following sums up the conclusions of the Society 

 based on the first investigation: 



Dosage. — For a child over two years old the dose of anti- 

 toxin should be, in all laryngeal cases with stenosis, and 

 in all other severe cases, 1500 to 2000 units for the first 

 injection, to be repeated in from eighteen to twenty-four 

 hours if there is no improvement; a third dose after a 

 similar interval, if necessary. For severe cases in child- 

 ren under two years, and for mild cases over that age, the 

 initial dose should be 1000 units, to be repeated as above 

 if necessary; a second dose is not usually required. The 

 dosage should always be estimated in antitoxin units, and 

 not of the amount of serum. 



Quality of Antitoxin. — The most concentrated strength 

 of an absolutely reliable preparation. 



Time of Administration. — Antitoxin should be admin- 

 istered as early as possible on a clinical diagnosis, not 

 waiting for a bacteriological culture. However late the 



