1914] OTIS— TRANSPIRATION OF WATER PLANTS 461 



small amount 



becomes 



a part of the cell sap. 



emersed 

 lo matter 



demands, they are never temporarily 



may 



excessive transpiration; nor are water plants ordinarily subjected 

 to as great extremes of temperature, relative humidity, and other 

 external factors as are land plants. It is to be concluded, then, that 

 the relations between absorption and transpiration stated above 



time 



hours. In the experiments which follow, however, ] 

 than this have been chosen, usually of about 12 hours. 



Apparatus 





The apparatus which was finally adopted consisted essentially 

 of nine large tanks supported in the water by two pontoons. Each 

 tank was equipped with a simple device for measuring the evapora- 

 tion from the tank. A recording thermometer was housed in a small 

 shelter about 2 m. from the pontoons. Relative humidity was 

 ascertained by means of a cog or "egg-beater" psychrometer (4), 

 which was in each case rested on a support about midway of the 

 vertical height of the growing plants. A barometer of the United 

 States Weather Bureau pattern, but graduated in the metric sys- 

 tem, was maintained on shore and read at intervals for the air 

 pressure. Centigrade thermometers for ascertaining the tempera- 

 ture of the water both within and outside the tanks, and graduates 

 of various capacities for measuring the amount of water evaporated 

 were employed. In the experiments conducted on the Huron 

 River at Ann Arbor, a Green's recording anemometer making elec- 

 trical contact every tenth mile was employed in addition to the 



ioned. The cups of the anemometer revolved 



men 



dm 



2 An excellent summary describing instruments adapted to the measurement of 

 evaporation from water surfaces is contained in C. F. M vrvix's Methods and appara- 

 tus for the study of evaporation. II. Instruments (9). 



