146 GROWTH. [CH. VI 



to the upper end of the plant must be regulated. As it 

 is difficult to tie a knot at exactly the desired place, the 

 following plan, fig. 27, should be followed. Let an 

 assistant hold the lever up at the desired angle ; pass the 

 string from the plant over the wire loop hanging from 

 the short end of the lever : pull it tight and secure it by a 

 folded piece of gummed paper; tViere will be time to 

 regulate the length before the gum dries, and when it is 

 dry it is fairly secure unless the experiment is made in a 

 greenhouse or in damp air. To make it thoroughly 

 secure one or more ligatures of fine silk may be added 

 above the paper. If after all the index is not at the 

 right height, the fulcrum must be slightly shifted up or 

 down the supporting rod. The cord from the plant to 

 the lever must be vertical, which can be insured by 

 shifting the position of the flower-pot. 



Take a Narcissus growing in a flower-pot, and having 

 a scape in active growth, — the flower being still a young 

 bud. Place the pot on a layer of damp sand at the 

 bottom of a galvanised iron cylinder so supported that a 

 flame can be placed under it when it is desired to expose 

 the plant to a higher temperature. The manipulation 

 necessary in starting the experiment will probably in- 

 terfere with the growth of the scape, so that it should be 

 allowed to grow for 3 or 4 hours without further disturb- 

 ance. A spirit lamp or very small gas-flame is then 

 lighted under the cylinder, and the temperature of the air 

 within the cylinder carefully watched for some little time. 

 After 2 or 3 hours — during which, if the temperature has 

 been raised by some 10°, the steps traced on the drum 



