METHODS OF WTiKK. 185 



ference is tliat this tube of potato-broth, which is only the tliircl remove from the 

 colony, contains a pure culture of the organism with which I started, but simple 

 observations of the tube, even when coiiplecl with a ver}' firm persuasion, do not 

 assin-f Die that such is the fact. I check the inference by making plate-cultures and 

 find in the tube either ia) onlv the original organism; (K) a mixture of two or more 

 organisms; (c) a pure culture of some wholl}' different organism, wliich entered 

 during one of the transfers as an accidental contamination and has crowded out the 

 original organism. 



(2) A plant is inoculated from a solid culture or fluid culture of a supposed para- 

 site, and becomes diseased. The inference is that the inoculated organism lias caused 

 the disease. I check this inference ])y making plate-eultiires from the interior of 

 the diseased tissues and find (a) great numbers of the inoculated organism in pure 

 cultiu'e and capable of again producing the disease, which I determine b\' actual 

 experiment; (16) a mixture of organisms; (t) some wholly different organism; (d) no 

 bacteria whatever. 



(3) Fermentation-tubes of cane-sugar bouillon inoculated with a supposedly pure 

 culture soon show clouding in the closed end, with an abundant production of gas 

 and acid. The inference is that these phenomena are due to the presence of this 

 particular organism. I check this inference by making plate-cultures and find (a) a 

 pure culture of the original organism ; (b ) only an intruder ; [c] a mixture of two 

 organisms, in which case both may break up the sugar in the manner described, or 

 only one of them. 



(4) Drops of fluid containing a supposedly pure culture are dried on sterile 

 cover-glasses and subsequenth' put into sterile beef-broth, which becomes clouded. 

 The inference is that the organism in question has resisted the drsdng. 1 check 

 this inference by making plate-cultures from the fluid and find (^7) a pure culture of 

 the right organism ; {b) a pure culture of some intruder. 



(5) The thermal death-point of an organism is tested b}- inoculating tubes of 

 beef-bouillon and exposing them to a given temperature in the manner already 

 described. Subsequently the bouillon clouds or does not cloud, as the case may be. 

 The inferences are that the organism is killed or is not killed b)- the exposure. The 

 first inference is checked b}- having at the same time inoculated other tubes of the 

 same bouillon, which have been kept at room-temperatures, and which (a) do not 

 cloud, showing either that the bouillon itself inhibits growth or that onh- dead 

 organisms were inserted, /. ^., those from too old a culture ; or (b) Avhich cloud readily, 

 showing that failure to grow in the exposed tulles is actually, as it was presump- 

 tii-ely, attributable to the temperature of the Avater-bath. I check the second infer- 

 ence b}- making poured plates from the clouded tubes and find (a) pure cultures of 

 the right organism ; (b) pure cultures of some intruder. 



f6j A plant, which we will designate as A, is subject in the field to a certain 

 disease, and this disease is readih' reproduced under experimental conditions, using 

 pure cultures of a given microorganism. A related plant, which we designate as B, 

 is subject in the field to a similar disease. A microscopic examination shows sim- 

 ilar lesions associated with a morphologicall)- similar organism, and Petri-dish 



