. II. HALES'S APPARATUS. 41 



which explain these facts in a simple and satisfactory man- 

 ner. It is easy to show that the ascension of mercury takes 

 place equally in two tubes prepared in the manner of Hales, 

 but differing from each other in this, that in one of them the 

 tube containing the ashes is open at the top, while in the 

 other it is closed. 



It must, however, be observed, that the experiment 

 would not be attended with this result if the column of 

 ashes were short, or the latter less up-heaped. With an 

 apparatus similar to that of Hales I made the following ob- 

 servations. I luted a leaden tube to the top of a glass one 

 containing the pulverised substance. By the aid of this I 

 could easily remove the air from above the ashes. At the 

 moment when the quicksilver began to rise I produced a 

 vacuum, and the mercurial column not only did not de- 

 scend, but it continued to ascend. It is, then, indubitable 

 that the ashes form, above the column of water, a wall or 

 partition which performs the office of a closed tube ; in fact 

 Hales's apparatus is a barometer. In another experiment, 

 at the moment when the mercury began to rise, I covered 

 the whole with a bell-glass, and produced a vacuum ; the 

 mercury instantly fell. I have obtained the same results by 

 substituting a stem with leaves for the tube filled with ashes. 

 If I introduce the upper part of this stem into a balloon, as 

 I remove the air the mercury continues to rise, but on the 

 contrary, if I form a vacuum over the mercurial reservoir, 

 it immediately falls. Hence, then, we conclude that what 

 Hales calls the force of aspiration (imbibition) is a simple 

 barometrical phenomenon. Whether the column of ashes 

 or the leaves and trunk of a tree form the upper closed part 

 of the barometer, the water penetrates the powders or vege- 

 table tissues by imbibition, and the atmospheric pressure 

 gradually effects the rise of the liquid. 



Exhalation by Leaves. We must, however, remark a 



