100 LOSS OF WEIGHT. [CH. IV 



weighed, with the bottle of water, 287 grams : by using 

 one of Becker's balances without a glass case, and having 

 a beam supported 40 cm. above the table of the balance, it 

 was possible to place the bottle on the pan and arrange 

 the leaves and branches so as to be clear of the scale-pan 

 knife-edges. 



Weighings and readings should be taken at hourly 

 intervals. The burette ought to be read to O'Ol c.c, if the 

 weight is recorded to 0"01 gram. It is instructive to 

 repeat the experiment with a freshly cut branch in which 

 the negative pressure is not satisfied. It will be seen 

 that the absorption is much greater than the loss by 

 weight, it may, for instance, be three times as great. 

 After taking a few hourly readings the apparatus should 

 be left to itself for 12 hours when equality between gain 

 and loss should be fairly established. 



(116) Spring Balance. 



We have found the following arrangement, fig. 19, useful 

 for demonstrating the loss of weight due to transpiration, 

 and it is probable that it may prove to be useful for 

 research purposes under certain conditions. A cut branch 

 in a test-tube of water T is suspended by a wire to a 

 spiral spring 8. At P the wire passes through a small 

 hole in a metal plate : at / a fine spun glass filament is 

 fastened horizontally to the wire. As T loses weight the 

 index rises and its movement is recorded by means of 

 a horizontal microscope. With the low power of our 

 microscope one degree of the ocular micrometer equals 

 0*044 mm.: the following readings (expressed in gradua- 



