280 SCHREINER AND REED: EXCRETIONS BY ROOTS 
natural gas and collecting the separated carbon on cool surfaces. 
It acts by absorbing part of the soluble matter from the soil ex- 
tract, a power which it possesses to a remarkable extent, by virtue 
of its enormous surface. The growth of plants in the extract 
after treatment with carbon black, ferric hydrate or other absorb- 
ing agents, is usually greatly increased. The conclusion logically 
follows that the retarded growth in the original soil extract is due 
to the presence of some substance or substances actually detri- 
mental to plant development and not to the absence of beneficial 
nutritive substances. 
This experiment is typical of a great number of experiments, 
employing various unproductive soils. In many cases the growth 
of plants was greatly improved by diluting the extract with dis- 
tilled water; in other cases by brief boiling, or by distillation, 
the toxic properties being found in the distillate. Without dwell- 
ing here at length upon the exact data derived from such experi- 
ments, it may be said that they agree in showing that the unpro- 
ductiveness of those soils was due to the presence of substances 
which exerted a toxic action upon plants. It was also shown that 
the toxic effect in the soil extracts could be overcome in various 
ways. 
Experiments upon the diminished yields of succeeding crops . 
have given results which indicate that the harmful effect of con- 
tinuous planting of the same crop may be due to the production 
of deleterious substances. Many, perhaps a majority, of investi- 
gators have assumed that the diminished yield of a second crop 
is the result of the depletion of the plant nutrients by the first 
crop. There is now evidence from a number of sources that an 
important factor in causing diminished yield is the presence of sub- 
stances detrimental to plant growth. An experiment giving evi- 
dence on this point has been described by Livingston (05) in 
which wheat was planted in a series of five pot cultures of clean 
glass sand, simultaneously with five other pot cultures planted in 
glass sand which had previously grown wheat for twenty-one 
days. The two series were subjected to the same conditions and 
growth was measured by the amount of water which the plants 
transpired. The growth of the plants in the ‘exhausted”’ sand 
was about 45 per cent. of that in fresh sand. In the same experi- 
