DETERMINATION BY TORSION BALANCE 205 



crease of weight of the living plant by the new method is 

 directly measured, and not after drying and powdering it, 

 a procedure which must introduce many errors. 



8. Periodic determination during insolation. — The fact 

 that the increase of weight is due to photosynthesis is 

 proved by periodic measurement of the increase of weight ; 

 the increase is found to depend on the intensity and 

 duration of light. 



9. Reduction of period of experiment. — Finally the in- 

 crease of weight can be demonstrated after an exposure of 

 the plant to light for as short a period as one hour. 



I will next describe in detail the method of weighing. 



Torsion Balance for Determination of the 

 Increased Weight 



A simple form of Torsion Balance, previously described, 

 was found to be sufficiently sensitive for the determina- 

 tion of the increase of weight. The two arms of the balance 

 are formed of light aluminium of thickness sufficient to 

 ensure rigidity. A very fine wire of phosphor-bronze, 

 used for galvanometer suspension, is passed through the 

 centre of the arms of the balance and kept taut by an 

 adjusting screw. The two arms are bent into L-shape to 

 throw the centre of gravity of the system below the point 

 of suspension. The right arm of the balance carries two 

 rings at its end, the upper one of which carries a rider (not 

 shown in the figure) of aluminium wire 3-5 mg. in weight. 

 This is used, as will be presently explained, for the purpose 

 of calibrating the value of the movement of the horizontal 

 index of the balance. The sprig of Hydrilla is suspended 

 in water by a hook from the lower of the two rings. The 

 suspending thread is a single cocoon-fibre, which is waxed 

 to prevent it being wetted. The thread is so fine that its 

 movement up or down in water produces no change in the 

 weight caused by slight difference in the length of immer- 

 sion. The left arm of the balance is prolonged into an 



