526 PHYSIOLOGY OF GROWTH. 



seedlings of Phaseolus multiflorus. If we grow these in loose 

 garden soil, and cut off the epicotyl, just above the soil, when it 

 has reached a length of a few cm., the buds present in the axils of 

 the cotyledons develop into shoots, in place of the organ removed, 

 and soon appear above-ground. 



It is a well-known fact that the stalks of the flower buds in 

 almost all species of the genus Papaver, at a certain stage of 

 development, curve downwards, and this, as far as is known, 

 depends on correlation. The bud, in organic connection with its 

 stalk, exerts on it a certain influence, which causes it to curve 

 strongly and bend downwards. Proof of this is afforded by the 

 following experiment, first performed by Vochting, 2 and which I 

 repeated with good results using a poppy plant growing in the 

 open. We cut a number of buds from their stalks, lay some of 

 them aside, and fasten others to their stalks again by means of fine 

 silk. The curvature of the stalks always disappears after a time, 

 and our experiment teaches that its occurrence is by no means 

 the result simply of the weight of the bud, but depends on corre- 

 lation. 



It is highly instructive, in connection with our present subject 

 to pay particular attention to the peculiarities and behaviour of the 

 bud scales of different plants. 3 In Aesculus and Pavia, the outer 

 scales of the winter buds are brown and membranous. Then come 

 succulent, green, very large scales, and finally the foliage leaves. 

 Investigation of the history of development teaches that all the bud 

 scales are merely foliage leaves arrested at an early stage of de- 

 velopment, and the following experiment leads to the same result. 

 If, just after the winter buds have sprouted, we cut off the tips of 

 Aesculus or Pavia shoots, and, without removing the shoots from 

 the parent plant, strip off their leaves, then in the course of the 

 summer the buds lying in the axils of the leaves develop into 

 foliage shoots, whereas normally, they would become winter buds. 

 The notable point about this case of correlation lies, however, in 

 the fact that these shoots produce no bud scales, but only (at least, 

 so I observed in my experiments) forms intermediate between scale 

 leaves and foliage leaves, together with foliage leaves. The lower 

 leaves of the shoot have small but still segmented lamina?, and 

 these are seated on a green scale-like leaf structure, while the 

 leaves at a higher level have the form of normal foliage leaves. 

 There is thus in Aesculus and Pavia, as also in other plants, a very 

 well marked correlation between the presence or absence of the 



