[Vol. 2 



518 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



be well briefly to mention the results obtained by some of the 

 earlier investigators working on the products excreted by the 

 leaves of plants. De Saussure (1804) found that leaves im- 

 mersed in distilled water soon lose a considerable amount of 

 substance, composed for the most part of alkaline salts. 

 Treviranus ('38) mentioned the results of various workers 

 who studied the incrustation of minerals on the surface of 

 leaves and found it to consist of calcium and silicon salts, 

 especially of calcium carbonate. Gaudichaud ( '48) and Payen 

 ( '48) both found that there is an alkaline excretion on certain 

 parts of the leaves of some plants, yet they disagreed as to 

 the extent of this phenomenon in nature. Sachs ( '62) ascer- 

 tained that drops of water on the leaves of certain plants soon 

 become alkaline, which he considered to be the result of an out- 

 ward diffusion of alkaline salts in the leaf. Volkens ('84) 

 studied the deposit of calcium carbonate found on the leaves 

 of various plants. Dandeno ( '02) made a comprehensive study 

 of the different phases of the subject. Among other things, he 

 determined that the alkaline substances extracted from leaves 

 by distilled water are largely potassium and calcium carbo- 

 nates and probably potassium oxalate. He further found that 

 the residue from the evaporation of dew drops, guttation 

 drops, and of water used in drenching the leaves is practi- 

 cally the same, and is similar to the calcareous deposit found 

 upon the leaves of certain plants. The above investigations 

 may therefore be considered as tending to substantiate the idea 

 of exosmosis from leaves. 



III. Methods of Experimentation 



The methods used for the electrolytic determination of 

 exosmosis were the same as those described in the writer's 

 paper referred to above. In that contribution (Merrill, '15) 

 some of the curves were plotted on the basis of the specific 

 conductivity. In the present paper, however, all curves are 

 plotted on the basis of the values of x on the Wheatstone 

 bridge when the resistance in the box is 9,110 ohms ; as these 

 values increase the specific conductivity also increases. In 

 order to have a basis of comparison between the values of x 



