ADDENDA. 139 



twisted tight, or by a very strong tightly stretched rubber band, which 

 is prevented by the groove from slipping down the cover. The outer 

 pot could better be of aluminum. Another method, which weighing 

 tests have shown to be the best of all for preventing evaporation from 

 the soil (for the rubber is not a perfect preventive), is this: Remove a 

 plant from its pot and plant in a glass jar as previously described ; the 

 earth should not come to within a half-inch or more of the top. Wrap 

 the stem of the plant from just below the ground to an inch above it 

 with two or three turns of rubber, and pour over the soil successive 

 layers of melted paraffine heated barely hot enough to melt it. After it 

 has hardened it will shrink away from the jar, but may here be remelted 

 by a hot iron. This method does not injure the plant, and seems a per- 

 fect preventive of loss of water otherwise than through the shoot. It 

 is well to cover the jar with black paper to shield the new young roots 

 from intense light. Amongst efficient balances for use in transpiration, 

 that figured in Pfefifer-Ewart, page 241, should be noted. 



5 (page 81). A decided improvement in the potometer is the fol- 

 lowing: Instead of turning the tube down at one end (Fig. 15 en the 

 right) it should be turned upward and connected with an upright tube 

 like that beside the shoot. The water placed in this tube supplies some 

 pressure to the shoot and appears to make it work more evenly than 

 when the water is drawn from the vial below the tube. Bubbles of air 

 can be admitted by a small hole bored through the tube (by the method 

 given on page 43), or by a break in the tube, kept covered by a piece of 

 rubber tubing which can be slipped aside to admit the air. The tube 

 must be perfectly clean inside, or the bubble of air will move unevenly. 



6 (page 83). An excellent method of applying the cobalt paper is 

 the following: With the largest cork-borer cut out rings from two good 

 corks, so that they will be 5 mm. high and thick. Glue pieces of long- 

 piled velvet or plush upon one face of each, and when dry trim it down 

 and remove with the cork-borer, so that it finally covers only the ring of 

 cork. Over the other face of the ring glue a cover-glass of proper 

 size. Make from elastic wire a spring-loop like those of test-tube hold- 

 ers, and push the two ends into the corks. The two corks should now be 

 held with their velvet faces against one another by the spring of 

 the wire. A piece of the cobalt paper is then placed in each one, 

 the apparatus is sprung open, and the corks are allowed to rest on the 

 two surfaces of the leaf: the velvet makes closed chambers against 

 even the inequalities of the leaf, and the change of color in the cobalt 

 paper may be clearly followed through the cover-glasses. A new form 

 of hygrometer, especially for use in determining loss of water from sto- 

 mata, is described by MacDougal in Torreya, I. 16. 



7 (page 89). I have obtained excellent results by the use of the 

 above-mentioned colored liquids placed in small square bottles glued side 



