84 JOHNKS DISEASE 



the cultures obtained from us. In liver broth to which 

 blood-serum and the glycerine extract mentioned 

 above had been added, Holth observed, after an inter- 

 val of about four weeks, a slight growth which 

 appeared as a sediment at the bottom of the tubes. 

 This growth gradually increased in amount, and later 

 formed a film on the surface of the medium. The 

 surface growth was often observed to be thick and 

 knobby. Attempts to subculture the bacillus on to 

 media not containing extracts of tubercle bacilli or 

 their dead -bodies were unsuccessful, with the possible 

 exception of serum-agar containing i per cent, of 

 somatose and i per cent, of Heyden's "Nahrstoff," on 

 both of which a slight growth might have been 

 detected. 



Regarding the addition of serum to the medium as 

 advocated by Holth, we found that it did not improve 

 our original egg medium, so performed no further 

 experiments on these lines until after the publication 

 of Holth's paper, when we tested dog's ascitic fluid 

 with liver broth-agar containing an extract of B. phlei. 

 The unheated ascitic fluid was added just before 

 setting the agar, and for one batch the fluid was 

 first passed through a Doulton porcelain filter. This 

 series of experiments showed the agar to be improved 

 by the addition of the fluid, especially when it was 

 filtered before being added to the agar. With the 

 addition of blood-serum, however, we did not obtain 

 such good results, but this may have been due to the 

 fact that we used rabbit serum, while Holth used 

 horse serum. 



Since Holth's paper, M'Fadyean, Sheather, and Ed- 

 wards have published similar results. These workers 

 isolated the bacillus from twenty-three naturally in- 

 fected cases, and from four experimentally inoculated 



