88 LOSS OF WEIGHT. [CH. IV 



well below the eosin solution which should not more than 

 half fill the bottle. When the air-pump is set in action, 

 it is obvious that its tendency is to suck out the contents 

 of the tracheids at the cut end, and as a fact air bubbles 

 are seen to issue at that point. Nevertheless, in spite of 

 this the eosin rises in the branch. Leave the pump running 

 for 6 or 7 hours, when the branch should be sawn off 

 above the cork, without stopping the pump, so as to avoid 

 injection of the wood with eosin. Readings of the baro- 

 meter should be taken in the course of the experiment 

 and compared with the readings of the pump-manometer'. 



Section B. Loss of Water by Transpiration. 



(113) Loss of weight. 



To get a general idea of the amount of loss due to 

 transpiration it is well to take a series of weighings of a 

 plant growing in a flower pot. Select a 

 plant^ with a large leaf surface, in a small 

 flower pot so that it may not be too heavy 

 for the balance'. In order to confine the 



\ / 



loss by evaporation to the plant, the sur- 

 face of the earth must be covered with a 

 divided disc of sheet-cork painted over '**' ' ^'^' 

 with wax mixture. The pot is wrapped in sheet india- 



1 We have only used a negative pressure of 50 cm., but Strasburger used 

 72 cm. 



^ Jerusalem artichoke or Chrysanthemum. 



' We use a French druggist's balance capable of carrying 4 or 5 

 kilograms, and of turning with 0-5 gram when loaded with 1000 

 grams in each pan; it is a useful form of balance for the purpose in 

 question. The beam, etc, being below in the box, and the pans there- 

 fore free to take a tall plant. 



