282 SCHREINER AND REED: EXCRETIONS BY ROOTS 
had observed small drops tp exude during the night from roots of 
Viola arvensis growing in pure sand in a transparent dish. He 
even stated that he had observed small fragments of material at 
the extremities of the roots of certain other plants, which he 
believed to have been exuded from the roots. His observations 
rest upon what must now be regarded as insufficient evidence and 
appear to have been made without any of the precautions neces- 
sary for a scientific experiment. 
The idea that roots excrete waste matters was promulgated by 
von Humboldt and by de Candolle and by them given prominence 
in explaining natural plant associations and crop rotations. It is 
interesting to note that de Candolle (’32, 3: 1480) stated his 
belief that the cockscomb and other noxious weeds injured the 
neighboring plants by some excretion from the roots. It is a 
matter of common knowledge that de Candolle used his theory of 
root excretions as a basis for explaining the benefits of crop rota- 
tions. He reasoned that the excreta from the roots remaining in 
the soil would be harmful if the soil were again planted to the 
same crop; but that, if a different species were planted, it would 
receive very little harm from the excreta of the previous crop, and 
even possibly might be benefited by them. Plenck (’94) and 
Macaire-Prinsep (°32) also endorsed the idea of root excretion. 
Macaire-Prinsep made an experiment from which he and others 
drew incorrect conclusions. By separating the roots of a plant 
into two groups, one of which was placed in a flask containing 
pure water and the other in a flask containing a solution of 
sodium chloride or lead acetate, he found that traces of the solute 
could be detected in the flask originally containing pure water. 
He believed that the roots in the solution had taken up dissolved 
matter which had been again actively exuded by the roots. His 
work was criticised by Braconnot (’39) and Unger (’36), who 
showed that his results were due to the capillary action of the 
roots, aided, no doubt, by the siphon which they formed. Walser 
(38) and Braconnot (’39) attempted to detect excretions by eX 
amining the medium in which plants of the Papaveraceae had 
grown. They believed that it would be possible to demonstrate 
the presence of opium-like bodies in the medium if the roots pro- 
duced excretions. Without recognizing the fundamental error of 
| 
