118 Alfred J. Etvart : 



The following table from the Melbourne Observatory records gives 

 the hygrometric conditions for the first day of the experiment. 



March 13th, 1912. 



With a spread of 14 feet diameter, the tree covered an area of 

 ground of 142,588 square centimetres. With a rate of evaporation of 

 2.19 millimetres per 9 hours, this would give a total loss from a free 

 surface of water of 31 litres, or 3.4 litres per hour. The estimated rate 

 for the whole tree from branch observations made at 10 a.m., 1.30 

 p.m. and 4.30 p.m., represented an average total of 2.2 litres per hour 

 during the same period. This is considerably less than the actual 

 amount absorbed, and less than the amount that would have evapor- 

 ated from a free surface of water covering the same area as the spread 

 of the tree. Since the leaflets of the cut branches were, however, in 

 all cases partially folded by the close of the experiment, it is possible 

 that the estimated rate of transpiraton was somewhat less than 

 actually occurred in the tree as a whole. 



In all experiments in which eosin is used to indicate the ascent 

 of water, the lateral diffusion of the dye makes it not altogether a 

 perfectly safe guide as to the exact path of the Avater current. For 

 this reason, copper sulphate was used as the poison to precede the 

 eosin. ^ 



When copper sulphate is added to a strong solution of eosin, the 

 greater part of the dye is precipitated, and hence it was thought that 

 the ascending eosin would be fixed in the walls of the wood vessels 

 by the copper sulphate impregnating them, and so largely or entirely 

 prevented from lateral diffusion. This was actually the case. An 

 examination of the wood showed the almost exclusive part played by 

 the wood vessels in the ascent of sap. No indication could be seen of 

 any connection of the medullary ray cells with the a-scending stream, 

 but as the copper sulphate had killed then before the eosin had 

 entered tlie stem, this fact affords no evidence one way or another. 



In addition, the presence of copper sulphate in the wood caused the 

 eosin to be held back to such an extent as to make it useless as a 

 measure of the movement of the sap. Thus when the tree was cut 

 down at the end of the third day, eosin was only perceptil)le in the 

 main trunk up to a height of 10 feet, and was entirely restricted to 

 the outer layers of wood 1 to IJ inches in depth at the base, and 

 tapering to a depth of ^ inch upwards. 



