470 



NA TURE 



[September 8, 1904 



Czapek (02) quotes the experiment of Brunchorst, who 

 found that a circular cut round the tip, not deep enough to 

 free the terniiinal part, has the same effect as amputation. 

 On the other hand, Nemec ' states that geotropism persists, 

 if the root-tip is cut half through by two opposite incisions 

 in different planes, so that the whole of the tissues are 

 divided, and yet the tip is not amputated. Thus four out 

 of five bean-roots treated in this way showed distinct 

 geotropism in 55 hours. This seems to me a striking result, 

 as showing that the shock of the operation is not exclusively 

 the decisive element. Nemec has, moreover, shown that if 

 geotropic curvature has begun on a normal root, a wound 

 interferes with the amount of after-effect, and that the 

 precise nature of the wound is not decisive, and this, as far 

 as it goes, confirms the assumption that two half-cuts would 

 produce as much shock as actual amputation. 



Czapek - finds that splitting a bean-root longitudinally 

 has the same effect as decapitation. This would mean that 

 decapitation produces its results by shock only, since in a 

 split root there is no removal of the tip. I think I was the 

 first to make use of the splitting of roots in this connection. 

 I wished to show ^ the incorrectness of Wiesner's view — 

 viz., that amputation prevents geotropism by checking 

 growth. In my experiments the split roots were greatly 

 checked in growth, but curved geotropically, behaving in 

 this respect quite differently from amputated specimens. 



Another striking bit of evidence on Czapek 's side of the 

 question * is the fact that I^upin roots from which " \ mm. 

 of the tip has been removed, and which, therefore, contain 

 no statoliths." show the remarkable homogentisin reaction 

 which he has convincingly proved to be a symptom of gravi- 

 perception. Czapek adds that the same is true of roots 

 from W'hich i mm. has been removed. It seems to me that 

 Nemec 's reply to this ^ is of value. He finds that the 

 root-cap in Lupin is variable in length, but always longer 

 than 5 mm. ; therefore, in the roots from which \ mm. 

 only was removed there should have been some statocyte 

 tissue remaining. Even after the removal of i mm. 

 - the root can, according to Nemec, rapidly form stato- 

 cytes, since the section is in the neighbourhood of the 

 calyptrogen.' 



Nemec suggests it to be conceivable that differences of 

 pressure in Czapek 's sense may give rise to the homo- 

 gentisin reaction, while the true act of graviperception is 

 confined to the statoplasts. This is no doubt possible, but 

 I confess that, if the homogentisin reaction can occur in 

 root-tips which have no statoliths I should consider it a 

 strong argument in favour of the view that pressure- 

 difference in Czapek's sense supplies the machinery of 

 perception in roots. Czapek also claims that his experi- 

 ments with bent-glass tubes (Czapek, 95) prove the gravi- 

 perceptive region of the root not to be confined to the 

 region of statoplasts, since if the root-cap alone is in 

 the vertical branch of the tube, geotropic curvature is 

 not excluded. Nemec (04) has attempted a rejoinder 10 

 this objection ; with what success readers must judge for 

 themselves. 



It will be seen that, in my opinion, the balance of evidence 

 is not fatal to the statolith theory. Czapek, who treats the 

 question in a broad and liberal spirit, is by no means in- 

 clined to deny that statoliths have a share in gravipercep- 

 tion ; all he claims to prove is that the statoplasts do not 

 supply the whole of the mechanism. It is not easy for an 

 upholder of the theory to allow this much in the present 

 stage of the controversy. The best way of testing the 

 theory is by comparing the distribution of geotropism with 

 that of statoliths ; and if we are to allow, in all cases which 

 are opposed to the statolith theory, that the stimulus depends 

 on pressure differences in Czapek's sense, we deprive our- 

 selves of the best means of proving the truth or falsehood 

 of our theory. Those who uphold the theory must have the 

 courage of their opinions and finally trust to the facts of 

 distribution. But further knowledge is necessarv before 

 such a judgment can fairly be made. 



1 Nimec (01, p. 19). 2 Cz.ipel;, (98, p. =o=) and (02, p. 118). 



3 F. Darwin (82). J Czapek (02, p. 46S). 5 Neroec (04, p. 53). 



'* He adds that the calyptroeen>may in this way have an indirect importance, 

 and Firtsch's belief that this tissue >vas the essential seat of graviperception 

 may be accounted for. 



Centrifugal Force. 

 Jost ' objects that plants on a centrifugal machine do 

 not behave as the theory would lead us to expect. Thus he 

 found that certain roots and seedlings showed geotropic 

 curvature, although the statoplasts were scattered through 

 the cell, not spread out on the cell-walls furthest from the 

 axis of rotation. Miss Pertz - and I have repeated some 

 of Jost's experiments, and have come to an opposite con- 

 clusion. We find that Setaria does not curve with a centri- 

 fugal force of less than 002 g., and this is about the limit 

 for visible displacement of the starch-grains. As the centri- 

 fugal force increases up to 004 g. we get slight amounts of 

 curvature and slight amounts of starch displacement. The 

 two phenomena cannot be accurately compared, but so much 

 is clear : that the result of Knight's experiment is not 

 destructive of the statolith theory, but, on the contrary, is 

 roughly in harmony with it. 



The result of an intermittent stimulus may seem to some 

 a difficulty. Jost ' produced geotropic curvature by placing 

 seedlings in the horizontal and vertical positions for alternate 

 periods of 35 minutes. With alternate periods of 50" hori- 

 zontal and 2' 30" vertical he sometimes failed to get a 

 geotropic curve, and exposures if less than 50" always failed. 

 It is commonly said that 15-25 minutes are needed for the 

 starch to fall on to horizontal cell-walls, and it may seem, 

 therefore, that in these experiments neither 35 minutes nor, 

 a fortiori, 50" could produce a change of position in the 

 statoliths, and that therefore the experiment is destructive 

 to the theory. But this would be a wrong conclusion, for, 

 according to my experience, the falling time of starch is 

 often less than 15 minutes ; and even if this were not so 

 there would be no difficulty in understanding the above 

 experiments, for, as Jost allows (he. cit.), and as Nemec 

 (02) has also pointed out, the statoplasts may stimulate the 

 cell without the occurrence of any visible displacement ; for 

 if the statoplasts do not fall over and spread out on the 

 horizontal walls there must be a column or heap of starch- 

 grains, the height of which equals the width of the cell, 

 resting on the lateral wall of the cell instead of, as in the 

 normal position, a shallower layer pressing on the basal 

 wall. Here we have plain conditions of differentiation 

 between the vertical and horizontal positions. 



The same considerations apply to the whole question of 

 what is known as the geotropic presentation time' — i.e., the 

 minima! period of horizontality needed to induce a geotropic 

 curvature. It has been said that the presentation time 

 corresponds with the time needed for the statoliths to fait 

 on to the horizontal walls of the sensitive cells. It seems 

 to me that we hardly have knowledge enough to be certain 

 of this coincidence, and since, as above pointed out, the 

 statoliths may begin to stimulate before they are visibly 

 displaced, the question is not one of much interest or de- 

 serving of special inquiry. 



Theoretical. 



Elfving's^ well-known experiment with grass haulms 

 shows that (in this instance) the action of the klinostat 

 depends, not on the prevention of all graviperception, but 

 on the equal distribution of stimulus." But other plants 

 react differently — that is to say, they do not exhibit increased 

 rectilinear growth on the klinostat. This can best be 

 accounted for, as Noll ' suggests, by the supposition that the 

 equally distributed stimulus tends to produce a simultaneous 

 increase and decrease of growth-rate on opposite sides of 

 the rotating plant.* We, therefore, get in an indirect wav 

 evidence in favour of what has not been directly proved — 

 namely, that in geotropic curvature the diminution of growth 



1 Jost (02). 



- Darwin and Pertz (04). By an oversight we omitted to give a reference 

 to Neniec's (02, p. 347) interesting reply to Jost's criticism. 



s Jost (02), p. 175. See also Czapek (98), p. 206 ; and Noll (00), p. 462. 



NO. 18 19, VOL. 70] 



Czapek (98), p, 



^ Elfving (84) proved that the pulvint of grass hauli 

 when kept in slow rotation on a klinostat. 



<> My experiments on the germination of Cucurbita demonstrate the same 

 point (Darwin and Acton, 94) Czapek (02, p. 469) shows that the homo 

 gentisin reaction occurs on the klinostat. 



7 Noll (92, p. 35). 



S We have shown (Darwin and Pertz. 04) that in Setarin the statolith! 

 undergo changes of position on the klinostat, 

 stimuli. See Heine (85), who briefly describes s 



1 length 



of 



