air-bubbles in the atmometer or in connection 

 witb the glass wool plugs. It is impossible to 

 detect bubbles in the atmometer except by 

 taking the whole apparatus to pieces, while a 

 bubble in the tube may sometimes become so 

 large as to endanger the continuity of the water 

 column. 



The device to be described below was de- 

 signed by Mr. Frederick Musch of New Haven 

 as a result of experience in operating a series 

 of porcelain atmometers for Dr. Norman 

 Taylor — instruments equipped with the Living- 

 ston-Thone mounting. Like the latter, 3 the 

 Musch mounting depends on a mercury valve 

 to prevent the absorption of water, but here 

 the resemblance ceases. The Musch device 

 (Fig. 1, A) consists primarily of a J-shaped 

 glass feed-tube, the straight end of which (a) 

 extends up through the stopper of the reservoir 

 while the curved end (c) passes down into the 

 reservoir, reaching to within about half an inch 

 of the bottom when in place. The atmometer 

 is of course attached to the upper end of the 

 tube, and a column of mercury (Hg) in the 

 curve at the bottom of the J forms the valve. 

 The short arm of the J is long enough to pre- 

 vent the mercury splashing or being forced out 

 of the tube into the bottle. A second important 

 feature consists in a short side-arm (b), 

 attached to the tube about an inch and a half 

 above the bottom of the J and extending out- 

 ward and upward. The outside width of the 

 J and the outside distance between arm and 

 tube are gauged to insure their easy insertion 

 through the neck of the reservoir bottle. The 

 side arm is attached at a point just high 

 enough above the curve of the J to permit the 

 latter to clear the bottom of the neck, while 



3 And like several others that had been de- 

 scribed previously: see references in Science, 52, 

 p. 86 (1920). 



—2— 



