34 



used 100 plants for each experiment, but soon I had to 

 diminish this number to 50 and even later to 25, because 

 I could not otherwise obtain a sufficient number of expé- 

 rimental objects. 



There is however a second way of avoiding this diffi- 

 culty or at least of minimizing the influence ofvariability 

 on accuracy, and this second way has been of more help 

 to me than increasing the number of expérimental objects. 

 The déviations from the average, hère taken together 

 under the heading individual variability are caused for 

 the greater part if not wholly by différences in the external 

 conditions, the purity of the atmosphère perhaps in the 

 first place. The amplitude of the variability would be 

 much smaller, if it should be possible to eliminate thèse. 

 Acting according to this line of thought I hâve been 

 fairly successful in limiting the variability, at least my 

 figures show much greater concordance than those of 

 Mrs. Kutten-Pekelharing, and this greater concor- 

 dance cannot be wholly ascribed to the larger number of 

 plants used. 



NUTATIONS. 



At the end of his discussion of the nutations of Avena 

 Eothert says: 'j „Eine erschôpfende Untersuchung der 

 Oscillationen lag nicht in meiner Absicht; ich habe die- 

 selben nur darum berûcksichtigen mûssen, weil sie bei 

 meinen heliotropischen Versuchen eine Quelle von Tau- 

 schungen bilden konnten und mich anfânglich auch 

 thatsâchlich irre fûhrten". Thèse w^ords — including the last 

 ones — I can adopt with regard to my geotropic experi- 

 ments and they account for the observations following hère. 



At first I attempted to note down as a positive resuit 

 every Just perceptible curvation." Proceeding in this way 



1) W. Rothert, 1894, p. 33. 



