26 BULLETIN 52, HAWAII EXPERIMENT STATION. 
show that ferric iron was completely precipitated while the solution 
was still strongly acid because a colorless filtrate could be obtained 
showing a pH value of about 4.4. No trace of iron could be detected 
by the sulphocyanate method in this colorless acid filtrate. 
Apparently ferric iron is unavailable to many plants on most soils 
since it is completely precipitated while the solution is still strongly 
acid and becomes available only when it is* reduced to the ferrous 
form. This would emphasize a hitherto neglected function of humus 
and organic matter in the soil. In Figure 5 is also given the titration 
curve for ferrous sulphate. This is not very accurate because oxida- 
tion took place during titration, but is chiefly of interest in showing 
that a strong test for soluble ferrous iron could be obtained when the. 
solution was decidedly alkaline. Although possibly modified by the 
presence of other ions, this fundamental difference between the 
solubilities of ferric and ferrous iron throws much light on the manner 
in which chlorosis is induced. 
Figure 5 explains very clearly why chlorosis is induced on the 
manganiferous Hawaiian soils. In soils containing an excess of 
manganese dioxid the iron is kept oxidized to the ferric form and, 
consequently, is not sufficiently available to the plants, at least to 
those susceptible to chlorosis. Any iron which is added to the soil 
is immediately rendered unavailable and the effect of such soil then is 
depression in the assimilation of iron by plants growing on them. 
This explanation also applies to the effect of manganese dioxid in 
nutrient solution. The explanation is not so simple in case of man- 
ganous sulphate. Deatrick (12) has shown that the brownish-black 
deposit which forms on the roots of plants that are growing in solu- 
tions containing manganous salts is a deposit of manganese dioxid. 
A like deposit very probably occurs also in the tissues of the plant, 
and, if so, naturally hinders the assimilation of iron, as previously 
described. 
In addition to explaining the manner in which chlorosis is induced 
on manganese soils, Figure 5 throws much light on Gile's work on 
lime-induced chlorosis. Chlorosis will not occur on calcareous soils 
in the presence of plenty of organic matter or of other material 
which is capable of furnishing a supply of ferrous iron notwithstand- 
ing the fact that oxidation of ferrous iron occurs readily and reduction 
of the ferric iron to the available ferrous form is difficult in strongly 
alkaline solutions. This explains why Gile (15) did not find chlorosis 
in pineapples when large amounts of calcium carbonate were added 
to a soil which was very rich in humus. Moreover, it explains why 
Gile and Carrero (20) found that rice becomes chlorotic in calcareous 
soils with ordinary percentages of water, but grows normally when 
the soil is submerged. Reducing conditions of course prevail in sub- 
merged soil and ferrous iron is then available to the plant. 
The chief problem remaining unsolved in connection with chlorosis 
is why on the same soil, either manganiferous or calcareous, some 
plants become chlorotic while others do not. The manner in which 
susceptible plants become chlorotic has been explained. Plants that 
are immune apparently obtain sufficient iron for their requirements, 
either because such requirements are very small or through some 
special relation of their roots to the soil. It is suggested that those 
who are interested in this problem grow both susceptible and immune 
