36 COMPOXJND ORGANS OP PLANTS. 



it is only necessary to interpose a trencli of gravel or sand between 

 them and the premises they are forbidden. How is this to be 

 accounted for ? Are we to suppose, as some have done, a sort 

 of prescience on the part of the vegetable ? On the contrary, 

 is it not all clearly explicable on the principle of endosmosis ? 

 There is always, in the forming and vitally active cells at the 

 extremities of the roots, a thicker fluid than the fluid in earth ; 

 the fluid in the earth is attracted by endosmosis through the 

 cell walls into the system of the plant, and becoming assimi- 

 lated, the newly formed cells of the roots necessarily take the 

 direction of the most fertile and favorable soil. 



Roots developing in the soil have a natural tendency to an 

 avoidance of the light, whilst the stem and leaves seem to seek 

 for the same. Hence hyacinth bulbs will grow much better 

 in water-glasses which are of a dark color, than in white 

 unoolored ones. So also when Dutrochet caused a misseltoe 

 seed to germinate on the inside of a window pane, it sent its 

 roots inwards towards the apartment; when on the outside of 

 the pane it did the same. Hence when seed is sown by 

 nature or the hand of art, however the seed may fall, yet in 

 germination the radicle so bends itself as to sink perpendicu- 

 larly into the soil, whilst the stem rises perpendicularly 

 from it. 



The force with which the radicle or root descends is very 

 considerable, and many attempts have been made to change its 

 obstinate tendency to burrow in the ground, but without effect. 

 We know not yet the cause of this invmcible tendency of the 

 radicle towards the eartli's centre. It has been thought that 

 the humidity which exists in greater abundance in the soil 

 exercises a sort of attraction on the radicle, but Duhamel has 

 shown that it is not so. He caused seeds to germinate between 



