August 13, 1908] 



NA TURE 



i57 



A sensitiser must be a depolariser, directly or indirectly. It 

 must be a reducing agent in the broad sense of the term, or 

 it must be changed into one by the action of light. In either 

 case tlie sensitiser is decomposed by the action of light 

 on the sensitive plate." In support of this position he 

 quotes Bothamley's observation that sensitisers act and a 

 developable image is produced on exposure when the plate 

 is immersed in a powerfully reducing solution, as proving 

 that the dye is not oxidised by exposure as Abney sug- 

 gested, and shows that the instability of the dye on ex- 

 posure to light is generally acknowledged as a necessary 

 condition for it to act as a sensitiser. This theory also 

 explains ."Vbney's experiment of exposing a collodion film 

 stained with cyanine to the spectrum, and then coating it 

 with a silver bromide emulsion and developing. The silver 

 bromide was developed to an image over the absorption 

 band of the cyanine, though the silver salt had never been 

 exposed to light. 



.\mong the general conclusions referring to this part 

 of the subject that the author gives in his sumining up, 

 the two following have perhaps not been indicated in this 

 abstract, namely : — (i) whether any substance is reduced 

 or oxidised by light depends on the depolariser, and {2) all 

 sensitisers are light sensitive, but the light sensitiveness 

 and sensitising power need not run strictly parallel. 

 Finally, he maintains that the electrochemical theory of 

 light, first proposed by Grotthuss, accovmts for all the 

 known facts concerning the action of sensitisers and the 

 action of light upon salts. 



In a third chapter (Journal of PJiysical Cltcniistry, June) 

 Mr. Bancroft deals in a similar way with the action of 

 clilorine and bromine on organic bodies, and the effect of 

 halogen carriers upon the reactions in darkness and in 

 light. He deals with benzene and toluene, and the 

 formation of addition and substitution derivatives, dis- 

 tinguishing in the case of toluene between the replacement 

 of hydrogen in the ring and in the side chain. After 

 summing up the known facts and showing wherein other 

 tiieories are deficient, he quotes the suggestion of Bruner 

 that chain substitution is due to the bromine molecule and 

 ring substitution to free bromine atoms, the preponderance 

 of ring substitution when the reaction takes place in nitro- 

 benzene as a solvent being " certainly connected with the 

 fact that bromine and a portion of the acid of the poly- 

 bromides is dissociated into ions " in this solvent. Mr. 

 Bancroft agrees with this theory in the main, but carries 

 it further. He assumes that as light of suitable wave- 

 length increases the conductivity of most gases, appar- 

 ently by forming ions, this dissociation will probably give 

 products (atoms or ions) one half of which will be positive 

 and one half negative — whatever meaning may be attached 

 to these words. If the dissociation products are ions, these 

 words will have their usual meaning. To this assumption 

 he adds four others that are partly new : — (2) That there 

 is a slight reversible reaction between bromine and the 

 positive gaseous ion, 



Br., Br+r:;Br5 + Bt+. 

 (3) That substitutions take place in the ring when the 

 negative gaseous ions are present in excess; otherwise in 

 the side chain. (4) That the addition products of benzene 

 are to be considered as analogous to the chain substitution 

 products of toluene. (5) That there is a reversible dis- 

 sociation of the halogen carriers with formation of so- 

 called gaseous halogen ions, and whether these ions are 

 positive or negative depends on the nature of the carrier. 



Granting these assumptions, at low temperatures and in 

 the dark, Br,.Br+ will be formed, and therefore an excess 

 of negative bromine ions and ring substitution. With 

 rising temperature the proportion of positive to negative 

 bromine ions increases, and sunlight will increase the dis- 

 sociation of bromine and of the Br,.Br+, until the ratio 

 of positive to negative ions approaches unity, and then 

 side-chain substitution takes place. With reference to the 

 fifth assumption, in chlorides of iron, antimonv, molyb- 

 denum, and aluminium, there is no doubt that chlorine is 

 the negative radical. In iodine monochloride, phosphorus 

 pentachloride, and sulphur chloride, there is evidence of a 

 tendency to form positive gaseous chlorine ions. The 

 ruithor remarks that " it would strengthen the argument 

 if it were possible to show why negative chlorine sub- 



NO. 2024, VOL. 78] 



stitutes in the ring and positive chlorine in the side chain, 

 but I do not see any explanation which can be carried 

 through." A weak point in the argument is the behaviour 

 of aluminium chloride, which Goldschmidt and Larsen 

 have found to behave exactly like stannic chloride as a 

 carrier, although no lower chloride of aluminium is known. 

 But, on the other hand, Turrcntine has found evidence 

 of its existence by the electrolysis of a chloride solution 

 using an aluminium anode, though the aluminous salt has 

 not been isolated. C. J. 



BAROMETRIC GRADIENT AND WIND 

 FORCE.' 



""PHE relation between the wind and the pressure dis- 

 tribution is a fundamental question of dynamical 

 meteorology. In qualitative form it is expressed by Buys 

 Ballot's well-known law that in the northern hemisphere 

 an observer standing with his back to the wind has the 

 region of lowest pressure on his left. The idea of a 

 quantitative relation between gradient and wind has been 

 made familiar to meteorologists by the classic researches 

 of Guldberg and Mohn. It must, however, be admitted 

 that a comparison of the wind velocity calculated from 

 the pressure gradient by the method given by these authors 

 with anemometer readings has not given entirely satis- 

 factory results. The discrepancies are generally attributed 

 to surface friction, but this quantity has not proved amen- 

 able to theoretical treatment, a difficulty which seems to 

 have barred the way for further progress along these lines. 



We should expect the effects of surface friction to de- 

 crease rapidly with altitude, so that the results of kite 

 and balloon ascents should yield a more suitable material 

 for a comparison of theory and observation than the ob- 

 servations at ground level do. In the report before us, Mr. 

 Gold has used the data accumulated by the Prussian 

 Aeronautical Observatory for such a comparison. From 

 the distribution of pressure he has computed the gradient 

 velocity over Berlin by the formula given by Guldberg 

 and Mohn for each day of the year 1905, and has set 

 beside it the observed direction and velocity of the wind 

 at 1000 metres and 2000 metres above sea-level. The 

 agreement at 1000 metres is surprisingly close. In the 

 preface Dr. W. N. Shaw says of it : — " The general 

 result of the investigation is, in my opinion, to confirm 

 the suggestion that the adjustment of wind velocity to 

 gradient is an automatic process which may be looked 

 upon as a primary meteorological law, the results of 

 which are more and more apparent as the conditions 

 are more and more free from disturbing causes, mechanical 

 or meteorological." 



Cases of discrepancy between observed and calculated 

 values may thus be regarded as exceptional, and their 

 special investigation promises an interesting field for 

 research. Tile observations of wind made at the surface 

 may perhaps also be utilised in this manner if, as Mr. 

 Gold suggests, we prepare a series of coefficients, ap- 

 plicable to the individual stations, which will enable us 

 to allow for tlie disturbing efYects of surface friction. 



For work on these lines a simple means of calculating 

 gradient velocities from pressure distribution is required, 

 and a considerable section of the report is accordingly 

 devoted to providing it. Tables have been constructed 

 showing the gradient velocities corresponding with different 

 distances between consecutive isobars on charts of certain 

 specified scales in different latitudes, or, if we prefer it, 

 we have in the frontispiece a convenient scale for reading 

 off the values from the map of the daily weather report 

 or the working chart of the Meteorological Office. 



The tables and scale give the gradient velocity on the 

 assumption that the isobars are straight, i.e. that the 

 path of the wind is a straight line. If this condition is 

 not fulfilled, a simple correction has to be applied from 

 a second set of tables (or scale of concentric circles) to 

 allow for the curvature of the path of the wind. 



There is one point of principle which requires to be 

 mentioned in this connection. In applying the correction, 



1 Report to the Director of the Meteorological Office nn the Calculation 

 of Wind Velocity from Pre^surr Distrihulioii and on the Variation of the 

 Meteorological Element with .Altitude. Ry Ernest Gold, f London 

 Wyman and Sons, for H.M. Stationery Office. Price as. 61L 



