1896 } TOXIC ACTION OF DISSOLVED SALTS 99 
chloride solution with that of a solution of dialyzed iron,” which 
contains no ferric ions inasmuch as potassium ferrocyanide (which 
is a test for ferric ions) produces no precipitate in the solution. 
The absence of chlorine ions in the solution is shown by the fact 
that silver nitrate produces no turbidity in such a solution. The 
tables show conclusively that ferric ions are much more poison- 
ous than are the complex ions containing ferric iron in the 
dialyzed iron solution. 
20. FERROUS SULFATE (FeSO, ). 
(Begun March 7, 12 M.; closed March 8, 12 M.) 
Concentration 
gm. mol. per liter Length Remarks 
Tat00 18.0™™ dead, purplish spotted 
ee ee 19 “ “é “ee oe 
astou 240 alive, tip somewhat discolored 
eae 22:0: * dead, discolore 
51200 34.0.7 alive, root-tip dead 
st eo a 29.0 bad “ “ “ 
Table 20 shows that for ferrous ions the concentration limit 
at which the seedlings will just survive is ;;},;) gram ion per 
liter. The same limit obtains in case of the nickel and cobalt 
ions (tables 21 to 24). Two nickel salts and also two cobalt 
salts were tested in order to show again that it is immaterial at 
this dilution whether the nitrates or the sulphates are used; in 
other words, that the toxic action of the solutions is solely due 
to the ions of the metals present. 
21. NICKEL suLFATE (NiSO,). 
(Begun March 7, 11 A.M.; closed March 8, I2 M.) 
Concentration 
2m. mol. per liter Length Remarks 
Beso 19.07" dead 
aes 22:5 “ epee 
stho0 4.6 alive 
ee a 31.0 de “ 
cn 
18 i ; 
acco: i Solution of dialyzed iron used in the experiments of table 19 was prepared 
Tding to the directions given in the British pharmacopeeia. 
