NYCTITROPISM 507 



would appear to coincide with the fourth possibihty mentioned above, and a 

 curving is only possible if the osmotic swelling on the convex side is greater, or 

 reaches its maximum value more rapidly than it does on the concave side. This, 

 as a matter of fact, was what Pfeffer assumed, and he held that nyctitropic 

 curvature resulted from changes in the two half-articulations, similar in character 

 but of unequal rapidity. As the turgidity on the concave side gradually in- 

 creases the conditions become favourable for that reverse curvature which, as 

 we have remarked, always takes place. 



This conception cannot, however, be held to be directly proved, since 

 Pfeffer' s results, obtained from experiments with half-articulations can by no 

 means be universally confirmed ; on the contrary, certain experiments have 

 been made (Schwendener, 1898 ; Jost, 1898) on the behaviour of half -articula- 

 tions which gaveresultsentirelycontradictory of Pfeffer' s observations ; when 

 darkened, the upper half-articulation expanded whilst the under one contracted. 

 This would tend to show that the alterations in the degree of illumination in- 

 fluenced the expansive efforts of the two halves in exactly opposite ways ; the 

 reverse curvature would not then be an immediate result of an external stimulus 

 as in Pfeffer's theory, but, as in tendrils and in nyctitropic growth movements, 

 would, in the first instance, be a result of the reaction arising from autotropism. 

 [WiEDERSHEiM (1904) is inclined to refer the dissimilar results of the removal 

 of half-articulations to differences in the mode of carrying out the experiments. 

 He maintains that his own special observations support Pfeffer's theory, but we 

 are not inclined to agree with him in that conclusion.] 



We must content ourselves at present with pointing out that a completely 

 satisfactory theory of nyctitropic pulvinus movements is not as yet forthcoming ; 

 such a theory can only be established after new and exhaustive experimental 

 research. For reasons which we will appreciate later, it will be necessary first 

 of all to study nyctitropic movement in leaves which have been previously 

 cultivated under perfectly constant external conditions, more especially where 

 there are no variations in light and temperature. Inquiry must then be made 

 whether complete extirpation of one half-articulation really moderates the reac- 

 tion of the other half in its original form, or whether it influences it correlatively. 

 Further, it is necessary to prove whether the cell-membranes remain entirely 

 unaltered so far as their elasticity is concerned, and whether the alterations in 

 expansion are due only to alterations of osmotic pressure. The increase in the 

 resistance to flexion of the entire articulation in the evening has not yet been 

 established beyond doubt, for Schwendener (1897) has recently been unable to 

 confirm it, and it did not always make its appearance in Brucke's experiments 

 (compare p. 506). 



Should further research actually establish the fact that the alterations in 

 the resistance to flexion of the articulation either do not occur regularly 

 or have not the significance which we have ascribed to them, then the separa- 

 tion of genuine nyctitropic movement from so-called day-sleep movement 

 cannot be carried out. In the case of Oxalis, Pfeffer (1876) found that the 

 day-sleep position in direct sunlight went hand in hand with a decrease in the 

 resistance to flexion of the articulation ; this decrease must have arisen from 

 a relaxation in both halves of the articulation, and especially in a greater relaxa- 

 tion in the concave half. Whether this relaxation is a phenomenon of general 

 occurrence in day-sleep and whether it has essential significance in this relation 

 must be left to further research to determine. At the same time, in all such 

 experiments efforts must be made to discriminate sleep from the heliotropic 

 profile position more accurately than is as yet possible ; at the same time it 

 cannot be denied that many profile positions are conditioned by a combination 

 of heliotropism and nyctitropism. 



While we regard such a combination of heliotropic and nyctitropic move- 



