414 BOTANICAL GAZETTE [JUNE 
action were those that gave unmistakable results with a majority of 
the tests. SCHERER’s test—the isolation of leucin and tyrosin— 
was of course conclusive. The application of this rule resulted in 
some cases in comparatively wide gaps betiveen the toxic and non- 
toxic strengths given in the table; e. g., m/650 Mg(NO,), is given 
as inhibiting, since it gave none of the tests; m/1000 as allowing 
action, since it gave all; while m/850 gave faint tryptophan reaction 
but no SCHERER’S test, and was therefore disregarded. 
In beginning to experiment with a particular salt, a stock solution 
was first made; a portion of this was diluted to give the greatest 
strengths desired, and greater dilutions were made from this by 
successive additions of distilled water. For example, in my first 
experiment with CuSO, a series of tubes with m/5000, m/ 10000, up 
to m/400000 were made up in duplicate. The results indicated 
where the critical points—between m/50000 and m/1000co—were 
to be expected, the duplicates preventing error. Closer series were 
then made and repeated until the limits of toxicity and non-toxicity 
were clearly determined. Finally the whole table was twice checked 
over in duplicate, so that the figures given represent in every case 
from eight to twelve concordant results. 
The results obtained may be shown by arranging the metals used 
in the order of toxicity, beginning with the most poisonous. The 
results obtained by Marnews, working with eggs of Fundulus 
heteroclitus, and those of McGurcan with diastase are given for 
purposes of comparison. 
MATHEWS (14) McGuican (16) CALDWELL 
Ag Ag 
Ag 
Hg Hg Hg 
Cu Cu Cu 
Cd Cd Pb 
Pb Co Zn 
Zn Zn Ba 
Co Pb Cd 
Li Sr Co 
Sr Ba Na 
Na Mg Li 
Ba is Sr 
Mg Na Mg 
NH, NH, 
Fi at a i 
