CH. vn] KNIGHT'S EXPERIMENT. 169 



that the geotropic parts of plants bend in relation to 

 centrifugal force. 



Another simple plan is to use a water-wheel driven by 

 a strong fine jet of water directed against the wheel from 

 the water-tap. The wheel should stand in a sink fitted 

 with a cover; in this way, with the help of the spray 

 from the wheel the experimental plants are kept 

 thoroughly damp. 



We use an apparatus designed by Mr H. Darwin. 

 A disc covered with a thick layer of cork is attached to a 

 horizontal axis turning on bicycle ball-bearings. It turns 

 with ease and is driven at considerable velocity by an 

 endless band from a turbine. The experimental plants 

 are kept damp by a bell-jar which is not attached to 

 the revolving disc, but fits by its broad ground edge 

 against a fixed vertical metal plate, through which the 

 axis passes. The space in which the plants rotate is 

 not therefore absolutely closed, but the air can be kept 

 sufficiently damp for practical purposes. The most serious 

 drawback to the apparatus is that the plants are 

 subjected to a current of air produced by their own 

 rotation. This evil has been fairly well overcome by a 

 four- armed fan attached to the disc, and dividing the 

 space inside the bell into four compartments ; as the fan 

 rotates, the air within the bell-jar is carried round with 

 the plants. 



To use the apparatus it is only necessary to pin 

 seedling beans so that the root lies tangentially; each 

 bean must be fixed on two pins firmly driven into the 

 cork. They should be fixed near the circumference of the 



