MOSAIC DISEASE OF TOBACCO. 107 



be used, otherwise there is a strong possibility, particularly in the case 

 of enzymes, that instead of a filtration occurring at once, a large amount 

 of certain constituents may be adsorbed (dependent on the nature of 

 the filter), and that true filtration may not take place until compara- 

 tively large amounts have been drawn through the filter. The writer 

 has noted this particularly in work with enzymes, many of which are 

 strongly adsorbed by various substances. Aside from the "ultramicro- 

 scopic" organisms, however, the bacteria cannot pass through many 

 of these filters. 



With reference to the causal agent in mosaic sap it has been found that 

 it passes through both the Chamberland and Berkefeld filters, and even 

 the finer grade of Berkefeld filter allows the passage of the causal agent. 

 Beijerinck {loc. cit.) showed that the juice was still infectious after being 

 passed through the Chamberland filter, and Allard {loc. cit.) and CUnton 

 {loc. cit.) have both shown that the juice was infectious after passage 

 through the Berkefeld (normal) filter. The results obtained by the 

 writer agree with these observations, and also the juice was found to be 

 infectious after passing it through the fine Berkefeld candle. The Kitasato 

 filter was also used, and here positive infection was also obtained, although 

 the percentage was small. The writer attempted to repeat these experi- 

 ments with the Kitasato filter during the past year, but was unable to 

 obtain the filter. In all cases relatively large amounts of the sap were 

 used after filtration through paper. 



The average percentage of infection obtained with each filter in the 

 writer's experiments was as follows: — 



Per Cent. 

 Chamberland (average of 3 examinations, 1911), . . . . .91.0 



Berkefeld (normal; average of 5 examinations, 1911), .... 03. 



Berkefeld (fine; one test only, 1914) 47.0 



Kitasato (average of 2 examinations, not dated) , . . . . . 40 . 5 



The work with the fine grade of Berkefeld and Kitasato filters should 

 be repeated, but there are sufficient indications to warrant the insertion 

 of these results at this time. 



Resistance to Antiseptics. — The writer has at various times treated 

 filtered and unfiltered juice with many of the antiseptics such as are 

 commonly used to prevent bacterial action. 



The following table contains the data and results obtained in one 

 typical series of experiments of this character: — 



