236 BOTANICAL GAZETTE [SEPTEMBER 
guaiac or potassium iodide directly was destroyed, but its peroxi- 
dase activity was still maintained, as shown by a vigorous action 
on the addition of hydrogen peroxide. 
The full significance of the change effected in the extract by 
treatment with the oxidizing agent was only apparent, however, 
after determining quantitatively its oxidizing ability and comparing 
it with the activity of the fresh extract in the presence of a constant 
supply of oxygen from hydrogen peroxide. This was done in the 
following manner. To toocc. of a solution consisting of equal 
volumes of 2M potassium ‘odide and 2M acetic acid 20 cc. of the 
treated horseradish extract was added and the amount of oxidation 
measured by titrating samples of 5 cc., at frequent intervals, with 
o.o1M Na.S.0;. The curve OA, fig. 2, expresses the results in 
such form that they may be compared with the peroxide reaction 
which has already been described and which is plotted in the curve 
OC of this figure. 
From this comparison it will be apparent that the amount of 
oxidation by the new preparation is small; but it was found possible 
after equilibrium had been nearly reached to renew its activity 
by a second treatment with potassium permanganate. For this 
purpose the ferment was removed from the potassium iodide solu- 
tion (after it had been acting 5 minutes) by adding about 2 volumes 
of 95 per cent alcohol. The small amount of white precipitate 
which appeared (about half an hour after adding the alcohol) -con- 
tained the greater part of the peroxidase. This was centrifuged out, 
washed in 80 per cent alcohol, dissolved in a few cc. of water, and 
again treated with potassiumi permanganate in precisely the same 
manner as before. On adding this preparation to a potassium | 
iodide solution as previously used, oxidation followed at a rate 
indicated by the curve AB (fig. 2). The smaller amount and 
somewhat slower rate of oxidation are probably due to loss of fer- 
ment in the precipitation and subsequent treatment. Anattempt to 
repeat the process a third time was not successful, only a faint cloudi- 
ness appearing on the addition of alcohol to the reaction mixture. 
By an inspection of the curves (fig. 2) it is apparent that if it 
were mechanically possible to repeat the process indefinitely the 
oxidation could probably be indefinitely prolonged. Moreover, by 
