begins. After that, the apex of the rhizome once more bends downwards into 

 the earth, then moves horizontally, and then again upwards. These alternations 

 of negative and positive geotropic curvatures with the transitional plagiotropic 

 positions show themselves to be related to the developmental state of the plant, 

 but they are, as we shall see by and by, partly dependent on external factors. 

 The rhizome of Paris is originally orthotropic, but when it once becomes plagio- 

 tropic (horizontal) it remains so if the external relations be unaltered. In the 

 majority of rhizomes the above-ground flowering shoot does not arise, as in 

 Adoxa or Paris, as a lateral outgrowth from the chief axis, but the end of the 

 chief axis itself comes above ground and becomes orthotropic, while a lateral 

 shoot goes on growing horizontally and continues the rhizome. This phenomenon 

 is seen e.g. in Heleocharis, Scirpus, Anemone nemorosa, and many other plants. 

 In all cases the chief axis grows horizontally in the first years until the flowering 

 period comes on, and, further, the chief axis of the seedling is always originally 

 orthotropic. The first alteration from orthotropy to plagiotropy is due to 

 unknown factors, but the annual or more frequently recurring (Heleocharis} 

 changes from plagiotropy to orthotropy and vice versa appear to be bound 

 up with alterations in the morphological and physiological characters of the 

 flowering shoot. 



Readjustments of a similar nature, dependent on the degree of maturity 

 of the organ in question, are known to be frequent in the flower axis ( VOCHTING, 

 1882). The flower bud of Agapanthus, for instance, is negatively geotropic, 

 the flower is horizontal, i.e. plagiotropic, the fruit is positively geotropic. 

 The poppy is another well-known example. Its buds exhibit positive geo- 

 tropic curvatures, which are compensated later by negative geotropism. WIES- 

 NER'S (1902) view that the nutation of the poppy bud was due to a special 

 ' weight curving ' is not well founded ; VOCHTING' s thesis, which has been 

 described above, has not been met by WIESNER ; especially is evidence wanting 

 that the nutation ceases when the weight of the flower is compensated. 



Let us now turn to cases where plagiotropism is obviously dependent on 

 influences of correlation. 



The influence of the chief axis on the lateral branches is markedly shown 

 when the apex of root or shoot is removed. It has been known for long that 

 when the apical bud of a spruce is removed the plagiotropic lateral shoots of the 

 uppermost whorl become erect and that the strongest of them becomes com- 

 pletely orthotropic and replaces the chief axis. A corresponding experiment 

 has been successfully carried out by SACHS on the root. It thus becomes a 

 pertinent question whether the plagiotropism of the lateral branches is a special 

 character inherent in the branch or whether the rest position in this case is the 

 resultant of two directing forces, so that, though naturally orthotropic, these 

 branches are diverted from the perpendicular by a force emanating from the chief 

 axis. This conception can scarcely be maintained in this form, for BARANETZKY 

 has recently (1901) shown conclusively that the geotropism of the lateral 

 branches is not essentially different from that of the chief axis, and that the 

 actual rest position is conditioned by a special property of the branch. More de- 

 tailed study will be necessary to appreciate properly what is signified by this 

 peculiarity which all organs possess and which we may term ' autotropism ' 

 (PFEFFER, 1893). 



VOCHTING (1882) was the first to show that when a shoot which had suf- 

 fered geotropic curvature was removed from the unilateral influence of gravity by 

 being placed on a klinostat, the plant attempted to compensate that curvature. 

 The concave side whose growth had been retarded during the curving now 

 elongated more vigorously than the convex side, and hence the shoot became 

 straight once more. It has been already shown that not only in geotropic 

 curvatures of the stem but generally in all cases of induced alteration of form 



