Chap. VIII. DIAHELIOTEOPISM. 439 



of diakeliotropism than that offered by many seed- 

 lings, the cotyledons of which are extended hori- 

 zontally. When they first burst from their seed-coats 

 they are in contact and stand in various positions, 

 often vertically upwards ; they soon diverge, and tliis 

 is effected by epinasty, which, as we have seen, is a 

 modified form of circumnutation. After they have 

 diverged to their full extent, they retain nearly the 

 same position, though brightly illuminated all day 

 long from above, with their lower surfaces close to the 

 ground and thus much shaded. There is therefore a 

 great contrast in the degree of illumination of their 

 upper and lower surfaces, and if they were heliotropic 

 they would bend quickly upwards. It must not, how- 

 ever, be supposed that such cotyledons are immovably 

 fixed in a horizontal position. When seedlings are 

 exposed before a window, their hypocotyls, which are 

 highly heliotropic, bend quickly towards it, and the 

 upper surfaces of their cotyledons still remain ex- 

 posed at right angles to the light ; but if the hypo- 

 cotyl is secured so that it cannot bend, the cotyledons 

 themselves change their position. If the two are 

 placed in the line of the entering light, the one 

 furthest from it rises up and that nearest to it often 

 sinks down ; if placed transversely to the light, they 

 twist a little laterally; so that in every case they 

 endeavour to place their upper surfaces at right angles 

 to the light. So it notoriously is with the leaves on 

 plants nailed against a wall, or grown in front of a 

 window. A moderate amount of light suffices to in- 

 duce such movements ; all that is necessary is that the 

 light should steadily strike the plants in an oblique 

 direction. With respect to the above twisting move- 

 ment of cotyledons, Frank has given many and much 

 more striking instances in the case of the leaves on 

 29 



