CH. VI] TEMPERATURE. 143 



recording surface. Take a Bath-Oliver biscuit tin, or any 

 other sufficiently tall cylindrical box : pierce two holes 

 just below the point to which the cover overlaps : pass a 

 stout wire through the holes and support the wire so that 

 the box may hang with its long axis vertical and its lower 

 end close to the table. To the lower end fix a cork with 

 sealing-wax, and into the cork thrust vertically a strong 

 short pin. Take a cheap American clock and fix it to the 

 table in such a way that the axis on which the hand 

 moves is vertical, while the hand itself moves in a horizontal 

 plane. The clock is so placed that the hour-hand just 

 touches the pin fixed to the biscuit box and at each hour 

 causes the box to oscillate on the supporting wire. 



(183) Temperature : microscopic method. 



Select a bean root about 20 mm. in length, impale it 

 on a pin and fix it to the cover of a flat-sided plate glass 

 vessel of water, so that the root is immersed and the hilum 

 just touches the surface. Having carefully levelled the 

 microscope, focus the tip of the root-cap, and note the tem- 

 perature of the water and the time of the observation. If 

 the root-cap is slimy and ragged, it may be gently cleaned in 

 the fingers. The reading will gain considerably in sharp- 

 ness if the root is illuminated horizontally by means of an 

 oblique mirror. Bean roots are so slightly heliotropic that 

 there is no objection to a lateral light. Other roots, e.g. 

 those of mustard, can only be observed when rotated on a 

 vertical axis. Spontaneous curvatures do however occur 

 in the bean root, and may spoil the experiment: the 

 curvature known as Sachs' curvature which occurs in the 



