428 Mr. D. Thoday. Experimental Researches on [Mar. 1, 



from the midrib showed the greatest tendency to flag, while towards the tip 

 the water supply appeared to be much more efficient. The results for 

 Leaf 17 illustrate this. 



Experiment 5. — August 7, 1909. Four leaves gathered 6.20 A.M., while still 

 dewy. Experimental half-leaves exposed, under a pressure of 1-^ metre of 

 water, at 7.45 A.M. Canvas screen, transmitting 40 per cent, of the light, 

 put up at 8 A.M. Experiment closed at 3.45 p.m. ; duration, 8 hours. Bright 

 sunshine throughout ; some haze. Maximum sun temperature, 27°'8 C. 



In the case of Leaves 19 and 22, two pieces were cut from each half-leaf 

 and treated separately as in Expt. 4. The stamped areas were in all cases 

 19"8 sq. cm. 



Table V. — Asymmetry errors as in Expt. 2. 



Leaf. 



Rate of increase in 

 milligrammes. 



Condition of experimental 

 half-leaf. 



T 19, tip 

 1 19, base 



20 



21 

 r 22, tip 

 \ 22, base 



16 -6 

 19-1 

 18 -6 

 9-8 

 9-7 

 10 -5 



1 Turgid. 



Stiffly turgid. 

 |- Turgid, then limp. 



Leaf 22 was found to have flagged somewhat at 11 A.M., and from that 

 time till the end of the experiment it remained a little limp, notwithstanding 

 the great head of water under which it was set up. It showed a correspond- 

 ing, rather low, rate of increase. 



The other three leaves all remained throughout the experiment completely 

 turgid. The high rates of increase shown by Leaves 19 and 20 are thus in 

 accordance with expectation. 



Leaf 21, on the other hand, showed a rate of increase comparable with 

 that of Leaf 22. The result was surprising, as the leaf was observed during 

 the experiment to be very stiffly erect, as if the head of water had freer 

 • play through its conducting channels than through those of any other leaf. 



Observations whicli had been made during the experiment with a horn 

 hygroscope of the form described by E. Darwin,* afforded no explanation of 

 the result in question. Readings were taken at intervals of about an hour on 

 both surfaces of each leaf, in the centre of each of the stamped areas. The 

 average of the readings obtained on tlio upper and lower sides were as 

 follows : — 



Leaf 19, tip, ; base, i^. Leaf 20, ff. Leaf 21, f^. 

 Leaf 22, tip, fj- ; base, JA. 



* " Observations on Stomata," ' I'liil. Ti iins.,' 1R98, B, vol. 190. 



9 



