410 BOTANICAL GAZETTE [JUNE 
It was hoped that the limits of toxicity might be more clearly defined 
in the case of enzymes. 
Very little work has been done in this field. In a general way it 
is known that slight amounts of metallic salts affect enzymic action. 
Cotes (6) found that cations and hydroxyl ions depress and anions 
accelerate the action of ptyalin and invertin. Results with acids 
and alkalis vary with the enzyme used (7, 8, 10). So far as I am 
able to ascertain, the only extended work with poisonous metals has 
been that of McGuican (16) upon diastase. His results confirm 
the theory advanced by MATHEWws (10, 11) that the physiological 
action of any atom or ion depends primarily upon its affinity for its 
electrical charge. The work described in this paper was undertaken 
with a view to determining whether this theory would hold for a 
tryptic enzyme, and was completed before MGGuican’s results were 
known to me. 
Bromelin was chosen as a typical vegetable trypsin, rapid in its 
action, easily prepared, and hitherto studied only with reference to 
its digestion-products and its action in the presence of acids and 
alkalis. 
: PREPARATION OF THE ENZYME. 
It was desirable to prepare the enzyme in as pure a state as pos- 
sible, since it was early found that the presence of associated proteid 
matter obscured the results of my experiments. Even very slight 
amounts apparently shielded the enzyme from the action of poisons. 
Furthermore, it was shown by some preliminary experiments, sum- 
marized elsewhere in this paper, that autodigestion, with the forma- 
tion of peptones, leucin, and tyrosin, occurred in solutions of the 
impure enzyme when kept for some time at temperatures between 
25 and 65° C. 
After several trials, a preparation was obtained which contained 
only very slight traces of associated proteid, and which was not at 
all autodigestive. The dialyzed NaCl-precipitate was dissolved in 
a little water, reprecipitated by the addition of 95 per cent. alcohol, 
again dissolved in water, and again precipitated by adding crystals 
of (NH,),SO,. A little of the associated proteid is left behind at 
every precipitation, and five repetitions of the process gave a prepa- 
ration containing only slight traces of proteid. Precipitation was 
much hastened by placing the vessel in the refrigerator at about 4° C- 
