New York AGRICULTURAL EXPERIMENT STATION. 141 
and sterile broths like (b) and solutions like (c). A detailed 
discussion of these points occurs later. 
Similarly sterile blocks from living carrot root were placed 
in the filtrate in comparison with like blocks in living cultures 
and others in broth plus 10% of chloroform to sterilize. The 
latter were shaken thoroughly and sterility proved. The car- 
rot tissues in the living cultures and in the tubes sterilized by 
chloroform were alike softened in three days, whereas the 
blocks in the filtrate were more slowly acted upon, requiring 
pine days for full action. Still further evidence of this rela- 
tion of filtration to enzym content was obtained by the method 
of alcoholic precipitation to be discussed later. Samples of 
filtered and unfiltered cultures were rendered 80% alcoholic 
and the filtered yielded only one-fourth ag much of the enzym- 
containing precipitate as did the unfiltered. Moreover five 
per ct. solutions of each of these tested upon razor sections of 
carrot and turnip showed the enzymic action of the unfiltered 
fully twice as rapid as that of the filtered. Taking into con- 
sideration both the relative amounts and the relative strengths 
of these solutions, it would seem from this last trial that pos- 
sibly four-fifths of the enzym was lost by filtration through 
the porcelain. 
Comparisons and conclusions.—All these experiments give 
like evidence that passage through the Pasteur-Chamberland 
filters as used in these trials reduces decidedly the enzym con- 
tent of the broth, although it does not remove all of it. Just 
why this retention of the enzym occurs has not been deter- 
mined. As ialready suggested, this may be in part, at least, 
the enzym contained within the bacterial cells and which would 
later diffuse into the surrounding liquid; or it may be in part 
or wholly external to these cells either closely associated with 
the bodies of the bacteria and so retained with the bacterial 
slime in the porcelain; or it may be that the filter removes 
some of the enzym content which is diffused or in solution 
in the broth. The results of Freudenreich (1899) point to 
the latter conclusion. He attempted to clarify cheese extract 
for qualitative analysis by passing through Chamberland fil- 
