Doyle — Some Researches in Experimental Morphology. 427 



go beyond our owu experiments for proof. Thus the leaf grafted on the 

 stem was also subjected to severe traumatic stimulus, but the secondary 

 thickening in it is by far less tlian that in the petiole upon which the sprout 

 was grafted. An observer of Plate XXXIII, fig. 3, might maintain, 

 however, that hei'e we have indeed a large wood development, perhaps due 

 to a wound stimulus. But, remembering Freundlich's (21) experiment in 

 severing the main nerve of a leaf in which he even got bridges of tracheides 

 growing round the cut to tlie lower end of the bundle, we fail to notice 

 any marked increase in tliat bundle either above or below the cut. Again, 

 Jost (28) describes an experiment in which he cut througli the bundle of one 

 of tlie primary leaves of Phaseolus multifjorus in a very yoiing plant. The 

 leaf grew well, but there was no increase in the bundle either above or 

 below the cut. Until, then, we get some positive evidence of the extent to 

 which a wound stimulus can increase the development of wood in a vascular 

 bundle, we can hardly admit that that stimulus has, at any rate, any great 

 direct importance in the secondary developments in the petiole. 



Loss of Correlational Influences. — To measure tlie effect of sucli a factor 

 as correlation seems futile, since we know so little of the real causes of plant 

 development. That it has a widespread effect we know well. The whole 

 phenomena of regeneration are bound up with it. Eemove the apical buds 

 of Bryophyllum, and the buds on the leaves develop. Remove the lamina 

 from the petiole, and the petiole soon dies. But it does not seem that the 

 mere correlational changes brought about by the suppression of the buds on 

 the parent plant have any direct effect on the large secondary development. 

 The same loss of correlational influences was a factor in those plants on which 

 only one leaf was left, the plant having been completely debudded. In this 

 D group there was less change than in any other. In the leaf cuttings, where all 

 correlational influences had been removed, the secondary vascular changes 

 were very slight. But that these influences have some powerful and 

 fundamental effect we cannot deny. But it is an indefinable indirect 

 effect. Thus, for the successful taking of the graft, all the parent buds must 

 be kept removed. It is the removal of the retarding influences exercised 

 by them which allows the so-called " permanent " tissue of the petiole to 

 react as it does to the otiier profound physiological stimuli which act upon 

 it after the grafting. It is like exploding a mine by releasing a spring. 

 The spring is no direct cause of the violent explosion, but without its release 

 no explosion occurs. Similarly the petiole is released from the correlational 

 influences (the spring), under the influences of the graft (i.e. the spark firing 

 the mine), it " explodes " into a very much thickened petiole. Only from 

 this point of view, I tliink, can we consider the correlational influences as 

 working at all. 



