Newcombe, Thigmotropism of terrestrial roots. 83 



brouglit against tliem, and lience were in contact ^vitll a con- 

 siderable portion of the elongating zone, but gave only uncer- 

 tain results, most roots growing straight. 



A rapid and gentle percussion against one side of the elon- 

 gating zone produced no cnrves. 



To bring pressure to bear on a 1 arger area of tlie root, loops 

 of paper, collodion membrane, rubber tissue, bristles and of 

 thread were liung over the tips of the roots, eacli loop attaclied 

 to an independent, light pendulnm, and several loops made to 

 pull on -the elongating zone of eacli root, wliile the roots were 

 immersed in water. The results cannot be said to demonstrate 

 thigmotropism, though in the case of Raphanus sativum the 

 number of positive curves was a good indication of such reaction. 



The preceding experiments having done no more than to 

 indicate a thigmotropic sensitiveness in the roots of some plants, 

 the next series of tests was made on roots in a damp-chaniber 

 with gravitation neutralized by revolution on a klinostat. Many 

 of the means employed to give pressure in the preceding exper- 

 iments were again used in these, and generally with more pro- 

 nounced results. It was found that glass-rods pressing against 

 the elongating zone brought some positive curves in nearly every 

 set of roots. There was a posibility however that the reactions 

 were hydrotropic, the glass-rod diffusing moisture against the 

 side of the tip only 2 to 3 mm beyond. This surmise was shown 

 to be partly tnue by the employment of various devices. Still, 

 when the root-tip was suraounded by a liood of thin paper, there 

 were many curves toward the glass-rods, and this result must prob- 

 ably be accepted as demonstrating thigmotrojiism. 



Finally, more uniform and convincing results were obtained 

 by using a Avater-stream to give the pressure, while the roots 

 were covered by very thin collodion tubes. The tubes when 

 tested showed that in the velocities of current used there could 

 be almost no filtration of water through the tubes, certainly not 

 enough to furnish a Stimulus for a rheotropic curve. The posi- 

 tive curve canie as well when the whole of the root was cov- 

 ered by a collodion tube as when but the apical 25 millimeters 

 were covered. 



The last series of experiments gives strong evidence for the 

 identity of rheotro])ism and thigmotropism. The 3 species of 

 plants wliich gave cui'ves in water-streams when their roots were 

 covered with collodion tiil)es are among those previously shown 

 to be most responsive rheotropicnlly. The curves in the covered 

 and the uncovered roots havo the same appearance and both 

 show approximately the same latent periocl, the curves in the 

 covered roots, as might l)e expected, following a little later those 

 in the uncovered roots. 



If now Ave assume the identity of rheotroi^ism and thigmo- 

 tropism, we may say, as was said f or rheotropism ^j that species 



1) N e wcombe, The Rlieotropism of Roots. (Bot. Gazette XXXIII. 183.) 



6* 



