150 



Scientific Proceedings, Royal Dublin Society. 



transpire, the water in this horizontal tube is withdrawn into the 

 bottle; and the rate of its withdrawal can 

 easily be determined by noting the time 

 occupied by the water in receding from the 

 open end of the tube to a given mark on it. 

 On reaching this mark, by gently pushing 

 the cork somewhat more firmly into the neck 

 of the bottle, the water can be again made 

 to reach the open end of the tube, and thus 

 a second determination can be made, and 

 so on. The objections to this form of ap- 

 paratus are (1) the comparatively few read- 

 ings that can be taken, owing to the fact 

 that, to ensure water- and air-tightness, the 

 rubber cork has to be pushed in nearly as far 

 as it will go at the outset ; and (2), if left for 

 even a comparatively short time, the supply 

 of water in the tube becomes exhausted, air 

 enters the bottle, and hence, before starting 

 fresh readings, it is necessary to remove the 

 cork, fill the bottle with water, and re-adjust. 1 



It occurred to me that both of these ob- 

 jections could be done away with by intro- 

 ducing a tapped thistle-funnel into the bottle, 

 by which water could be let in to take the 

 place of that removed by the shoot, and which 

 could, at the same time, serve as a tempor- 

 ary reservoir during the time in which no 

 observations were being made, so that no 

 air should enter the bottle. 2 



In its final form, my potometer worked 

 itself out into that shown in the accom- 

 panying figure (fig. 1). A calcium chloride 

 tower is provided at the top with a double- 



1 Hall, Annals of Botany, 1901, p. 558, describes and figures a somewhat similar 

 form of apparatus, but with a three -holed cork, the third hole being provided with a 

 wooden rod which can be gradually pushed into the bottle so as to force the water back 

 into the capillary tube. 



2 Since this paper was read I find that Farmer has also recently made use of the 

 tapped thistle-funnel for this purpose. Ref. in Bot. Centralblatt, Dec. 1st, 1903, p. 535. 



CUP 



Fig. 1. 



