98 



NA TURE 



[July 22, 1909 



intensity of blue light at Washington on the average day 

 seems to indicate that there is a substantial amount of 

 scattering by small solid or liquid particles. 



The figures for Mount Wilson give us confidence in the 

 trustworthiness of Mr. Abbot's determination of the solar 

 constant, because it is clear that the total effect of the 

 atmosphere can be eliminated with greater certainty if it 

 is mainly due to the permanent constituents of the atmo- 

 sphere, and not to matter which is variable in amount. 

 As the top of Mount Wilson is less than iSoo metres above 

 sea-level, we may conclude that at high elevation the blue 

 colour of the sky is completely accounted for by molecular 

 scattering. 



That the scattering sets a limit to the transparency of 

 gases must be kept in mind in discussing problems of 

 solar and stellar physics. We may feel confident, for 

 example, that what is called the reversing layer can only 

 have a small thickness, for otherwise we should not be 

 able to observe so far into the ultra-violet as we do. 



The scattering may profoundly modify the character of 

 the spectrum, as I have explained in a paper on " Radia 

 tion through a Foggy Atmosphere " iAsirophysical Journal, 

 vol. xxi., p. i), in which it is shown how increased thick- 

 ness, without change of temperature, may convert an 

 absorption line into a bright line. It seems to me prob- 

 able that the bright and dark flocculi shown in spectro- 

 hcliograph pictures may find their explanation chiefly in 

 variation of thickness in the absorbing layer, the bright 

 patches being due to increased thickness. 



The transparency of interstellar space has recently re- 

 ceived a much needed critical discussion at the hands of 

 astronomers, and Prof. Turner has applied the law of 

 scattering to explain certain discrepancies between visual 

 and photographic magnitudes. The value he gives for the 

 opacity allows us to calculate the average density of the 

 matter which is diffused through space on the supposi- 

 tion that it is gaseous. If the value of (/t — i)/D (where 

 fi is the refractive index and D the density) be taken to 

 be approximately equal to that of air, I find that the 

 number of molecules per cubic centimetre in space would 

 have to be of the order of a million, and the mean free 

 path of the order of 3000 kilometres. 



Although not directly connected with the subject which 

 forms the main part of this communication, I would like 

 to point out that the same analysis which gives the 

 coefficient of extinction in terms of the retardation of 

 phase at the source of the scattering also gives a resultant 

 force acting on the molecule in the direction in which the 

 light is passing. When summed up for all the molecules 

 this force is found to be identical with what is generally 

 called the " pressure of light," for if E represents the 

 energy density, the force acting per unit volume on the 

 scattering molecules is found to be kE, where fc is the 

 roeflficient of extinction. 



There is a widespread impression that light pressure 

 only acts on particles the linear dimensions of which 

 include several wave-lengths of light, but this is not 

 correct. The determining factor is the extinction of light, 

 whether by scattering or by absorption, as indeed appears 

 if we take the view adopted in Prof. Poynting's work on 

 the subject that a propagation of momentum accompanies 

 the transmission of light. The momentum is destroyed 

 equally whether the molecules act as scattering or as 

 absorbing centres. The extinction by scattering near the 

 surface of stellar bodies does not, however, appear to be 

 sufficient to cause pny measurable effects comparable with 

 their gravitation. Arthur Schuster. 



The Fixation of Nitrogen by Soil Bacteria. 



May I be allowed through the columns of Nature to ask 

 Prof. W. B. Bottomley a few questions with regard to his 

 paper on " Some Effects of Nitrogen-fixing Bacteria on the 

 Growth of Non-leguminous Plants " (Proc. Roy. Soc, 

 B, Ixxxi., 1909, 2S7), abstracted in Nature of May 13 (vol. 

 Ixxx., p. 327), as I had not the opportunity of being present 

 when the paper was read? 



Prof. Bottomley bases his conclusions on experiments to 

 show that Pseudomonas, the bacterium associated with the 

 leguminous plants, will fix mere nitrogen in an artificial 

 NO. 2073, VOL. 81] 



culture, when Azotobacter is also present, than when alone. 

 He quotes the following results : — 



Control o'48 mgm. N. per 100 c.c. culture solution 



Pseudomonas alone 0'9I „ ,, ,, 



Pseudomonas and 



Azotobacter ... f24 ,, ,, ,, 



These difference.* would almost seem to be within the 

 range of experimental error, but in any case, does not the 

 demonstration require a further statement of how much 

 nitrogen Azotobacter alone would fix? Other observers are 

 accustomed to get fixations by Azotobacter alone of from 

 5 to 20 mgm. of nitrogen per 100 c.c. of such a culture 

 solution, the maximum being about 10 mgm. of nitrogen 

 fixed per gram of mannite. The only conclusion that could 

 be drawn from Prof. Bottomley 's figures would be that 

 Pseudomonas injuriously affects the power of Azotobacter 

 to fix nitrogen, supposing that a reasonably active culture 

 of the latter had been used. 



Turning to the field experiments, in which Prof. Bottom- 

 ley claims to get an increase of crop by adding cultures of 

 Pseudomonas and Azotobacter to soil which presumably 

 already contains both organisms, data for estimating the 

 probable experimental error are again lacking. From the 

 Rothamsted experiments, where we may assume the con- 

 ditions are more than usually favourable to exactitude, the 

 mean error of a pair of similarly treated plots in a single 

 year is about ±10 per cent., which would more than cover 

 the differences observed by Prof. Bottomley 's experiment 

 with oats. 



In another experiment with barley. Prof. Bottomley 

 obtained a higher percentage of nitrogen in the corn from 

 the treated strip than in the corn from the rest of the field, 

 i'76 against 1-55 per cent. (I presume that " milligrammes 

 of nitrogen per cent." is a clerical error). In view of the 

 comparatively small changes in the composition of the grain 

 of a cereal which are usually effected by large variations in 

 the supply of nutrient, I should like to know from Prof. 

 Bottomley if duplicate samples were taken from different 

 parts of the untreated section of the field, and what range 

 of variation they showed in their nitrogen content. 



Coming to the next experiment, it is difficult to judge 

 how far a bulbous plant like Galtonia candicans is suitable 

 for experiments on nutrition, but it is rather necessary to 

 know what relation the weight of the bulbs planted bore 

 to those harvested. Prof. Bottomley only says that 250 

 bulbs " of equal size " were planted in each bed. Can 

 he let us have the weights in each case? Moreover, he 

 tells us that the treated bed was twice watered with the 

 culture solution, the control bed being given pure water a* 

 the same time ; was the same amount of water given to 

 each, and how much of the culture solution was applied 

 for it contained monopotassium phosphate, sodium chloride, 

 &-c., which may well have been a considerable factor in any 

 beneficial effect experienced? 



Prof. Bottomley will perhaps forgive me if these questions 

 may seem somewhat critical of his conclusions, but any 

 communication appearing in the Proceedings of the Royal 

 Society must be taken into account, and one therefore 

 wishes to have the data necessary for determining the 

 weight to be attached to the results. A. D. Hall. 



The Rothamsted Experimental Station, July 12, 1909. 



Occasional Unexplained Ringing of House-bells. 



An observation sent me by Mr. Alexander Sinclair, of 

 Swansea, to the effect that during a thunderstorm drops 

 of water leaking through the ceiling " assumed a pear 

 shape and jumped 9 inches almost horizontally to the 

 curtain rings above the window," suggests that house-bells 

 of the ordinary non-electric type may occasionally be rung 

 by this means. I picture the process as follows : — The bell 

 wires collect atmospheric electricity, by induction or other- 

 wise, which the walls are insufficiently conducting to carry 

 off freely ; consequently the bells get charged, are attracted 

 to a neighbouring wall or pipe, and released suddenly by 

 a spark. This little lateral jerk rings the bell. 



I put the simple suggestion on record because I sometimes 

 hear of an inclination to attribute the phenomenon to less 

 familiar causes. Oliver Lodge. 



