i88 



NA TURE 



'August 12, 1909 



of animals, to investigate life-histories of various 

 insects, parasites, &c., and generally to advise on 

 subjects relating to economic biology, agricultural 

 chemistry, and bacteriology. 



The articles in the journal are mainly summaries of 

 work done elsewhere rather than accounts of original 

 work; perhaps this was only to be expected from an 

 almost new laboratory. Mr. Collinge deals with the 

 use of lime, with special reference to its influence 

 on plant diseases like potato-scab and finger-and- 

 toe fungus ; he has also collected a good deal of 

 scattered work on the woolly aphis. Mr. Barlow 

 deals on similar lines with the effect on plants of 

 copper salts used as fungicides. The summaries 

 themselves call for no special comment, but the journal 

 as a whole is well got up. We shall be interested to 

 see how Sir Richard Cooper's experiment works — 

 whether the laboratory can maintain the detached 

 position essential for the publication of scientific work, 

 or whether, as has happened elsewhere, it becomes 

 merged in the purely commercial side. 

 Cambridge Comity Geographies : Somerset. By 



Francis A. Knight, assisted by Louie M. (Knight) 



Button. Pp. xi-l-192. (Cambridge: University 



Press, 1909.) Price is. 6d. 

 The characteristics of the series to which this volume 

 belongs were enumerated in our issue for May 13 

 (vol. Ixxx., p. 305), and much of what was written on 

 that occasion applies to the present book. The authors' 

 interpretation of tiie scope of geography is wide 

 enough to include a history of the county, its anti- 

 quities — ecclesiastical, military, and domestic — its ad- 

 ministration and roll of honour. Like previous 

 volumes in the series it is well illustrated, brightly 

 written, and generally attractive. 



LETTERS TO THE EDITOR. 

 [The Editor does not hold himself responsible for opinions 

 expressed by his correspondents. Neither can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or afiy other part of Nature. 

 No notice is taken of anonymous communications.] 



Difference between Longitudinal and Transversal 



Zeeman Effects in Helium Lines. 

 It is easily shown on the theory of electrons that the 

 amount of separation of the outer components of a trans- 

 versal Zeeman triplet must be slightly different from that 

 in a longitudinal doublet. Some time ago I showed 

 that the transversal separation in weak magnetic fields 

 does not strictly follow the linear relation with the 

 magnetising force, but, owing to an indirect method of 

 measurement, the exact amount of the separation could 

 not be measured with accuracy. By measuring the longi- 

 tudinal effect of helium lines with an echelon spectroscope of 

 thirty-five plates, each of i cm. thickness, made by Hilger, 

 I found that doublets can be distinctly separated in a field 

 of 180 gauss, when the right- and left-handed circu- 

 larly polarised light is linearly polarised in mutually 

 perpendicular directions, by interposing Fresnel's rhomb 

 in the course of the beam. Taking a number of points at 

 intervals of about 300 gauss from H = o to H = 20oo, and 

 ten to thirteen points from H = 20oo to H = 14,000, I found 

 that for the three lines AX = 6678, 5876, 5016, the relation 

 between the amount of separation 5\ and the strength 

 of the field H is exactly linear, so that 5X/H = constant 

 also in weak fields. In these experiments it was necessary 

 to gauge the strength of the field accurately for each 

 point before and after each micrometric measurement by 

 means of a small coil. The values of e/m were found to 

 be for 



X = 6678 (^/ot = I -86x107 



= 5876(03) = I -68x10' 



= 5016 =i-Sox 10' 



The separation of the satellite of D, is complex, but there 



is one component which gives the same value of e/m 



as D3. 



NO. 2076, VOL. Si] 



With the transversal effect the ratio 5A/H is not constant 

 in weak fields. With D3, the curve representing the rela- 

 tion between H and S\ is such that it increases very 

 slowly to H = Soo, then rapidly to an infle.xion point in 

 H=:i7oo, makes a bend, and from H = 20oo follows an 

 accurately straight course up to H = 14,000, which is the 

 strongest field used in the present experiment. In the 

 latter part of the curve d(SX)/dH = constant, which is 

 smaller for the transversal than for the longitudinal effect, 

 so that the curves representing these effects cross each 

 other in H = i20o and H = 10,900. The initial part of the 

 curve for the transversal effect shows a striking resemblance 

 to that of magnetisation in ferromagnetic substances. 

 The satellite accompanying D3 shows remarkably complex 

 separation, as shown by Lohmann, but there are two com- 

 ponents which take a similar course to the principal line 

 D3. The lines are already separated before reaching the 

 inflexion point above mentioned, so that the method which 

 I used in my former experiments, is confined only to 

 weak fields. With the line 6678, the initial course of the 

 curve for transversal effect is similar to that of Dj, but 

 the inflexion point is reached in a higher field H = 27oo, 

 and the curve becomes a straight line from 11 = 3609 

 upwards. The curve for longitudinal effect lies entirely 

 above that for the transversal, and d{S\)ldH in strong 

 fields is greater for the former than for the latter. 



The usual calculation of e/tn is made on the supposi- 

 tion that SX/H = constant, which is strictly obeyed in the 

 longitudinal, but not in the transversal, effect ; the dis- 

 crepancy in the value of e/m calculated from longitudinal 

 and transversal effects is at once explained. The initial 

 course of the curve can be accounted for by Voigt's theory, 

 but the appearance of the inflexion point before attain- 

 ing the straight course presents some difficulty. The 

 resemblance of the curve of transversal effect to that of 

 magnetisation seems to have an important bearing on the 

 exposition of the theory, which would explain these 

 characteristic features. The extension of these experi- 

 ments to stronger fields and with different elements is 

 being undertaken. H. Nagaoka. 



Physical Institute, University of Tokyo, July 16. 



Natural Selection and Plant Evolution. 



The letter from Mr. James B. Johnston in Nature of 

 August 5 touches on many important points, which cannot 

 be fully dealt with in a letter of reasonable length. 



In his opening sentence the writer, speaking of 

 chapter xii. in " Darwin and Modern Science," says that 

 " there, perhaps for the first time, the evidence of the 

 fossils with regard to the influence of natural selection 

 has been fairly tackled " ; I may point out that the 

 chapter cited really relates mainly to evolution, and 

 especially phylogeny ; only the last section refers to 

 natural selection, a subject on which, from the nature of 

 the case, the fossil record can throw comparatively little 

 light. 



I cannot think that, on the main question, there can 

 be any very fundamental difference between the writer's 

 views and my own, for he says : — " In the case of the 

 Tertiary mammals the action of natural selection can be 

 very clearly demonstrated in numberless cases." Mr. 

 Johnston cannot seriously mean that he accepts natural 

 selection for animals and rejects it for plants. The ques- 

 tion is simply one of evidence. As I have myself pointed 

 out, the direct evidence for the derivation of one species 

 from another is at present less satisfactory in the plant 

 than in the animal record (" Darwin and Modern Science," 

 p. 204) ; on this point we may hope for new light from 

 further research, though, as regards the efficacy of natural 

 selection (an essentially different question), I doubt if 

 palaeontological evidence will ever be really decisive. 



My point in speaking of the evolution of the pollen-tube 

 and seed was to show that such characters are adaptive, 

 a view to which Mr. Johnston is not likely to object. In 

 the present position of biological science evidence of adapta- 

 tion is commonly accepted as presumptive evidence of the 

 action of natural selection. 



The question whether a belief in the efficacy of natural 

 selection can be regarded as " barring out all design from 

 the world in which we live " is not one that can be dis- 



