MERCURIAL GAUGES. 251 



variations of the pressure exerted by the sap of such species 

 as are noted for the abundance of their flow, consisted of a 

 syphon-tube of thick glass, the two legs' of which were eight 

 feet long, and about four inches apart. This was inverted 

 and attached to a support of inch board, on the centre of 

 which was fastened a scale divided to tenths of an inch. To 

 one leg of the tube at the top was adjusted a brass stopcock, 

 by means of small rubber hose, and to the stopcock was con- 

 nected by a brass coupling a piece of thick lead pipe of small 

 bore and convenient length, which was joined by another 

 stopcock to the trunk, root, or branch which was to be 

 tested. The stopcocks were so made, with a tube on the top, 

 that communication could be opened between the free air and 

 either the lead or the glass tubing at pleasure, and, when 

 closed from the air, the passage was open between the 

 mercury in the syphon-tube, the water in the lead pipe and 

 the sap in the tree. The object of this three-way cock was 

 to facilitate filling the tubes with water and mercury, and 

 allowing the escape of any gas which might find its way into 

 the apparatus from the tree. A sufficient quantity of mercury 

 was poured into the inverted syphon to fill the two legs to the 

 height of about forty inches, and the remainder of the leg 

 connected with the tree, as well as the lead pipe, was care- 

 fully filled with water, all air being excluded. The other leg 

 of the syphon-tube was left open to the atmosphere. When 

 the sap exerted a pressure, it was indicated by a depression of 

 the mercury in the closed leg of the glass tube and a rise in 

 the open end, the difference between the two columns showing 

 the pressure in inches of mercury. Suction into the tree was 

 marked by the rise of the mercury in the closed leg and its 

 depression in the open one, and in making the record the 

 minus sign was prefixed to the figures expressing the number 

 of inches of mercury. 



One of the difficulties encountered in these experiments 

 arose from the liability to leakage, either around the stopcock 

 inserted into the tree, or from accidental wounds to the bark 

 or small branches. In cases where the pressure was very 

 great, it was sometimes necessary to solder a heavy sheet of 

 lead to the stopcock and nail it to the tree with a packing of 

 white lead in oil. Much trouble was also experienced from 



