August 5, 1909J 



NA TURE 



159 



may recur slightly before or after the beginning of a 

 second round ; in either case there is a false suggestion 

 as to the rotation. The railings would make discon- 

 tinuous vision of the spokes of the motor wheels, and a 

 spoke might be seen upright in one gap but at slightly 

 different angles at other gaps. I do not feel that the 

 solution of the problem lies in this direction. 



Winchester College. W. B. Croft. 



Natural Selection and Plant Evolution. 



M.iNV readers of X.\TURi; must have beeji browsing with 

 delight in the goodly volume on " Darwin and Modern 

 Science " which Prof. Seward, of Cambridge, has taken 

 such admirable pains to collect. Of all its many chapters 

 few are more significant than that on the palaeonlological 

 plant record by Dr. D. H. Scott, because there, perhaps 

 for the first time, the evidence of the fossils with regard 

 to the influence of natural selection has been fairly tackled 

 by competent hands. 



Dr. 1). H. Scott does not attempt to maintain that the 

 record to-day is nearly so imperfect as it was when 

 Darwin wrote his famous chapter thereanent, fifty years 

 ago. Dr. Scott's namesake and collaborator from 

 Princeton speaks even of the record as, in some parts, 

 "crowded with embarrassing wealth of material"; and 

 yet what about evidence of natural selection? The present 

 writer ventured to say {Conlcinp. Rev., July, igo2, 

 p. 83) : — " Modern palaeobotanisfs furnish us with next to 

 no evidence at all of the work of Natural Selection in 

 evolving new species." Prof. Seward vehemently 

 challenged my statements ne.xt month ; yet, seven years 

 later, Dr. Scott feels constrained to tell us : — " As re- 

 gards direct evidence for the derivation of one species 

 from another there has probably been little advance since 

 Darw^in wrote." 



To put it more plainly, Dr. Scott is forced to admit 

 that he can adduce absolutely no satisfactory evidence at 

 all. All he does is to affirm his own firm conviction (as 

 it is Prof. Seward's too) that natural selection must have 

 been the chief agent ; and he instances two cases where 

 he thinks the possible inference extremely plausible, viz. 

 (i) the case of the pollen tube, quite absent in the 

 Palaeozoic seed-plants, found very short and imperfect in 

 the living cycads and ginkgos, and fully developed in 

 the angiosperms, but fossil proof of linking forms there 

 seems none ; (2) the embryo in the angiospermous seeds, 

 whilst Pala30zoic seeds contain none. It may, as he says, 

 be " impossible " to some " to resist the conclusion " 

 that the nursing of the embryo by the seed was a process 

 of adaptation. But, at any rate, there is no fossil proof 

 thereof ; and yet, as Dr. Scott will scarcely deny, there 

 surely ought to have been some hint and trace thereof, 

 the record being so comparatively rich and full as it is. 

 In the case of the Tertiary mammals the action of natural 

 selection can be very clearly demonstrated in numberless 

 cases. If natural selection was the factor in plant evolu- 

 tion too, why is the record so obstinately silent? 



Dr. Scott, like Prof. Seward, takes refuge in the thought 

 that our plant record, for many purposes, begins far too 

 late. " An immense proportion of the evolutionary history 

 ■lies behind the lowest fossiliferous rocks." My chief 

 object in writing this letter is to ask. Is there any valid 

 proof of this in regard to land plants, the matter specially 

 in hand? Their record begins, actually, in the Upper 

 Silurian, and though it is very, very meagre and imper- 

 fect, the traces are too widespread to be denied. To deny 

 the existence of known Upper Silurian plants is rankest 

 scepticism, though Dr. Scott avoids all reference to them 

 whatsoever. Why, even so very cautious an investigator 

 as Mr. Robert Kidston tells us of " a plant showing 

 woody structure," a plant so high as that, in the Lanark- 

 shire Ludlow beds (Summary of Progress of Geol. Survey 

 for 1897, p. 74). The most important Upper Silurian 

 plant-remains are probably those from the Tanne Grey- 

 wacke of the Harz, a fairly numerous and well-developed 

 series, of as^e a good deal lower than Wenlock. Drs. 

 Scott and Seward (" Encyclop. Brit.," Supplement) wish 

 to pronounce all these fern-like and other plants Devonian, 

 because of their facies ; but Sir Archibald Geikie ("Text- 

 book of Gcoloffv." ii.. p. 07^1 tells us that these Tanne 



NO. 2075, VOL. 81] 



plants are found a long way below shales with graptolites, 

 which surely is proof enough of Silurian age. 



We have, then, fair evidence of land plants in the Upper 

 Silurian. Our very first air-breather or land animal, a 

 cockroach, conies from tl.e top of the Lower Silurian ; and 

 the fossil record of the whole Silurian is rich, varied, wide- 

 spread, without gap. Yet it yields no hint of conditions 

 favourable to land life below the top of the Lower 

 Silurian. Is it, then, scientific to postulate dogmatically 

 land plants earlier than the Silurian, simply because a 

 theory requires it? Dr. Scott admits quite freely that the 

 known facts go the other way. 



He tells us not only of the opposition of the mighty 

 like Niigeli, he also tells us that, as regards the suc- 

 cession of species, there are no greater living authori- 

 ties than Grand'Eury and Zeiller, and that, in their 

 opinion, " the evidence from continuous deposits favours 

 a somewhat sudden change from one specific form to 

 another." This is most certainly true of the palsonto- 

 logical record as a whole. The evidence is pverwhelming 

 here, if only our men of science would be brave enough 

 to forget their theories for a little while. Why insist on 

 exalting the a priori methods of the schoolmen on the fair 

 field of modern science? Why insist .on refusing all 

 evidence that does not suit? Why? Surely it is not, and 

 cannot be, to enjoy the pleasure of barring out all design 

 from the world in which we dwell. 



J.1MES B. Johnston. 



St. -Andrew's Manse, Falkirk. 



Musical Sands. 



I c.VNNOT call to mind the occasion upon which Dr. 

 Irving suggested that grains of hyaline quartz might pi^o- 

 duce the notes from musical sands, but, as ^a matter of 

 fact, the grains do not " ring," or vibrate 'individually as 

 sonorous bodies, and there is no apparent resqnance'or 

 sensible continuance of the note a'fter' the plunger is with- 

 drawn. I do not think any particular variet_y of silica 

 is essential, because coral-sand is often musical, and my 

 artificial musical sands are made up almost entirely of 

 silicate of iron. . ' 



I have already shown that the natural sorting actipn 

 of winds and waves' is a requisite condition for the forma- 

 tion of musical patches on sea-beaches. 



Mr. M. S. Gray's letter in Nature of July 29 giving 

 interesting particulars of his visit to the musical sand- 

 hill near Copiapo, in Chile, confirnis the statements made 

 by the inhabitants to Darwin in 1835. In his " Journal 

 during the Voyage of H.M.S. Beagle " Darwin referred 

 to this hill of sand as " El Bramador " — " the roarer or 

 bellower, " but he did not personally visit the spot. 



The extraordinary sensations experienced by Mr. Gray 

 were "probably similar to those which have been described 

 by the various writers who, from time to time, visited 

 Rig-i-RawAn and Jebel Nakous, both of which were re- 

 ferred to at length in my paper on musical sands in 1888. 

 Particulars of the artificial production of notes from 

 certain sands were also given bv me in Nature of August 

 6, 1S91. 



Cecil Carus-Wilson. 



A Question of Percentages. 



If a student obtains 37 out of 50 in one paper, 50 or 

 full marks in a second, and 71 out of 100 in another, 

 what is his percentage on the three taken together? If 

 we add the marks as they stand we get 158 out of 200, 

 or 79 per cent. If, on the other hand, we double the 

 marks on the first two papers, we have 74 per cent., 

 100 per cent., and 71 per cent. If we add these we get 

 245 out of 300, or 8i§ per cent. Will any of your mathe- 

 matical readers kindly tell me which of these two different 

 results is the true percentage for the three papers taken 

 together? The answer may be very obvious; I can see 

 that the two results must be different, but I cannot see 

 which is the more correct method to use. 



J. T. Cunningham. 



60 Milton Park, Highgate, N., July 24. 



