— 132 — 



The cotyledons of the Sorghums were as before cautiously 

 bent and fixed to cork plates. But in the present series the cork 

 plate was horizontal and so arranged that the h'ght Struck the 

 cotyledons more or less horizontally. Each preparation was placed 

 on wet sawdust and covered with a crystalising jar which was 

 darkened with black cah'co except at a narrow interval by which 

 h'ght entered. The jars were placed near a north window where 

 the temperature was between 20° and 30^' C, and the plants were 

 exposed to electric light at night when a long exposure was ne- 

 cessary. 



Position 1. (See fig. 4, in which the upper arrow gives the 

 line of light.) 



Ten experiments; in one case no curvature resulted, in the 

 other nine the hypocotyls curved from 20'^ to 70'^ in direction C, i. e., 

 in direction due to the Stimulation of the cotyledon. 



Position 2. (See fig. 5, in which the upper arrow gives the 

 line of light.) 



Fourteen experiments: in one no curvature, in three cases 

 curvature //, in ten cases curvature C. 



Thus in both positions the curvature was on the whole in the 

 direction C, i. e., that due to Stimulus of the cotyledon. It would 

 appear from the comparison of these experiments with the corre- 

 sponding ones on geotropism that the plants are more sensitive 

 to light than gravity, since heliotropic experiments in position 2 

 are more sucessful than with geotropism. We had some other 

 reason to suspect that light-grown Sorghums are rather sluggish 

 to gravitation. 



Heliotropic experiments, witli ghiss tubes. 



When the forcible curvature is produced by pushing the coty- 

 ledons into bent glass tubes we again found a strong tendency 

 to curve in the direction of the artificial bend in the cotyledon. 

 Thus in five experiments made with position 2 (Fig. 5), the cur- 

 vature was in every case in direction //; and since in the case of 

 heliotropism we know that photo-perception is practically confined 

 to the cotyledon we must believe that the y/-curvatures are traumatic. 



IV. Other Traumatic Curvatures. 



The following experiments were made to decide whether 

 injuring one side of the cotyledon induces curvature in the hypocotyl. 



