64 N e w c o m b e , Tliigmotropism of terrestrial roots. 



some of tliis filter-paper v.-as washecl in streng liydroclilorie acid, 

 then in water, then in 5 " q potassium liy droxide , again in water, 

 boiled in water for 15 minntes, and tlien allowed to lie in riin- 

 ning water for a montli. At tlie end of this period, tliis paper 

 was as effective as before in producing negative curves. The 

 same resnlt foUowed also whetlier the paper was last washed in 

 tap -water or distilled water. 



It is possible that bodies little or not at all hygroscopic 

 such as giass and niica niay not have tlie same same eftect as 

 the paper already used. To test this, a clean glass tnbe of small 

 diameter was broken into bits, and fragments not more than 

 1.5 mm in diameter attached by their concave surface to the 

 ti])8 of roots of Fisum sativmn , Bra.s6ica alba and Fagopyrum 

 esculenfum. Negative curves came promptly. When howewer 

 mica was used instead of the glass, all roots grew straight. 

 TliereujDon the glass fragments were washed in strong hydro- 

 chloric acid, in water, in ammonium hydroxide, and finally in 

 water. They were then heated in a dry oven to 150 " C. for 

 several minutes. AVhen these fragments were attached as be- 

 fore to the root-tips in a damp-chamber no responses were shown. 

 The former curves must have been caused by injurious n^aterial 

 on the glass. As noted on a preceding page, Detlef sen found 

 scales of glass so thin as to adhere closely to the curving sur- 

 face of the root-tip capable of producing negative curves. In 

 my ow]i work one such test was made with 7 seedlings of Fisum 

 .saf'iv'iw/, the giass scale covering about half of the cone of 

 the root-ti]:). No curves were formed. Inasmuch as Det- 

 lef sen found the cells beneath the giass scales dead, it is prob- 

 able that there was some injurious substance on the glass. 

 Fragments of hard porcelain and white marble were also used 

 in my experiments without effect in producing curves in the 

 i'oots of the three plants just mentioned. It is true that some 

 roots curved negatively, but some curved positively, and the 

 most of them did not curve at all. For instance: 59 seedlings 

 of Faf/o})t/rum esculentum had bits of giass attached to their 

 root-tips. Ten roots beut negatively, 3 positively and 46 re- 

 mained straight. 



To test the effect of objects having considerable ability to 

 imbibe water, yet not as much as the papers used before, thin 

 layers of plaster of Paris were cast on glass-plates, and, after 

 hardening, were cut up into Squares of 2 mm diameter. These 

 were used on the root-tips of F/sinn safivmn and Lup'iniis alhii^. 

 A total of 67 seedlings was employed. Twenty-three roots bent 

 negatively, and the other 44 remained straight. Similar results 

 Avere obtained with chips from clay flower-pots, and with un- 

 baked clay. Nine roots of Vicia faba L. had each a droplet of 

 plaster of Paris mixed with water placed on one side of the 

 root-tip where the plaster soon hardened. This is the experi- 

 ment of Nemec^j in which he obtained positive curves, and in- 



