27 



soil, but on the follouinf,^ clay number 2 had received three times 

 as much water as number i, and probably on account of sub- 

 irrigation was able to utilize a greater percentage of what was 

 given to it. 



Experiments with Covillca gave even more striking results. 

 September 5, the transpiration of two plants, designated i and 2, 

 was determined in the forenoon between i i and i 2, and again in 

 the afternoon between 3 and 4 o'clock. Number 2 was given 

 three ounces of water at 12:20, none being given to number i. 

 At the time of the afternoon observation it was found that num- 

 ber 2 was transpiring more than three times as rapidly as it was 

 before the water was given to it, and number i, which was not 

 watered, was transpiring only one-fifth as rapidly as it was in the 

 forenoon. 



Observations were also made for the purpose of ascertaining 

 the effect of exposure to direct sunlight in conjunction with water 

 supply. It was found that exposure to bright sunlight was uni- 

 forml)- followed by accelerated transpiration, whenever the plant 

 under observation had a full supply of water, but that otherwise 

 such acceleration did not take place. 



It is noteworthy that plants which had all along received a 

 meagre supply of water were nevertheless in a position to tran- 

 spire rapidly when once a full supply of water was furnished 

 them, while plants which from the beginning had received a very 

 large amount of water showed promptly a marked lowering in 

 rate of transpiration when the water supply was reduced. 



With so complicated a problem general statements may well 

 be made with extreme caution, but the evidence in the present 

 case is sufficient to show that in studies of transpiration it is alto- 

 gether unsafe to attempt to estimate an)- other factors whatever 

 without taking due account of water in the soil. 



Desert Bot.\nical Lahoratory of the Carnegie iNsriTiTifiN, 

 Tucson, Arizona. 



