IV.] VARIOUS THEORIES, &c. 121 



reservoir of water under the pressure of one atmo- 

 sphere, and if we then create a vacuum in the 

 mesophyll, it would not work. The water would at 

 length sink till its upper level was about ten meters 

 high, because under no conditions could the atmo- 

 spheric pressure support a higher column. In fact if 

 Boehm's system would work it would furnish a case 

 of " perpetual motion." 



Now take Hartig's theory. It may be said to 

 depend on the following facts. The conducting wood 

 always contains liquid water as well as water of 

 imbibition : the alburnum often has more water in the 

 upper parts than in the lower ;" whenever the amount 

 of water in the tree diminishes, the air-spaces in the 

 crown enlarge more than those in the stem, especially 

 below. Hence, the pressure of the air is less in the 

 upper parts of the tree. Especially striking are 

 Hartig's results with ringed trees, where some species 

 began to droop when the upper parts of the stem still 

 contained 70 per cent, and more of water : in these 

 cases it cannot be because no water, or too little, was 

 present, that the tree drooped, and the only alter- 

 native is that the conditions for driving the water up 

 were absent. As we have seen, Hartig assumes these 

 conditions to be the difference of pressure of the air- 

 bubbles the ringing lets in air, and the continued 



