September 28, 191 1] 



NATURE 



435 



The sectional meeting room was at first situated in 



Fawcett Road, a considerable distance from the other 



sections ; this position was found to be so inconvenient that 



a move was made to the Municipal College. It would 



appear important that Sections A, 13, and G be housed in 



oximity, so as to facilitate exchange of views 



members of these sections. The section was 



ed by several foreign chemists of distinction. 



The proceedings were opened with Prof. J. Walker's 



address, following which Prof. Carl Barus, of Providence 



University, U.S.A., read a paper on the diffusion of gases 



through water in which a novel and simple experimental 



Method was described. The method consists in finding the 



Emperature-pressure conditions for the flotation at a given 



hical cartesian diver, in which the gas to 



imprisoned. Slides were exhibited 



showing the details of the apparatus and the mass-time 



graphs for the interdiffusion of gases through water. 



•he graphs for a single gas are linear, those for 



are of indefinite variety, and the very curious 



result of a gas apparently diffusing against the pressure 



is frequently met with. Such anomalous results 



are explained in terms of the partial pressures of the 



constituents of the imprisoned impure gas. 



Dr. W. C. McC. Lewis dealt with the compressibility of 

 mercury. He regards the usually accepted value as in- 

 correct, since for mercury the difference between the latent 

 heat of vaporisation as determined and calculated from a 

 formula involving the compressibility is large. Dr. Lewis 

 of error in the experimental 

 ■ei nation : — 

 fi'l Effects produced by a Liver of absorbed air or 

 , or both, between the mercury and the walls of 

 the vessel; 



(2) That Ihe liquid (molasses mixture) into which the 

 "piston was dipped before insertion in the mercury is not 



ttely removed ; and 



(3) The unavoidable slip of the mercury past the piston. 

 All these effects act in the same direction ; i.e. they give 



rise to too great a volume decrease, that is, to too high 

 values of the compressibility. 



Dr. J. F. Thorpe followed with a somewhat technical 



the chemistry of the glutaconic acids, in which he 



experimental evidence to show that the molecule 



of glutaconic acid must have a symmetrical structure, and 



that the cause of the identity of the a and y positions must 



be of the same order as that determining the equality of 



1 positions in the benzene ring. Some interesting 



conclusions based on the experimental study of the pro- 



of the alkylglutaconic acids were also described. 



In a discussion. Dr. Lowry differed from the author's 



interpretation of his experiments. 



Mr. G. Le Bas gave a summary of an elaborate paper on 

 of constitutive influences on the molecular volumes 

 of organic compounds at the boiling point ; the subject is 

 ilex to allow of a brief abstract. 

 The last paper, by Prof. R. Wegscheider, dealt with the 

 of substituents on reaction velocities. The elucida- 

 tion of the laws governing the transformation of organic 

 fcmoounds is to lie found in the study of reaction velocities. 

 Prof. Wegscheider has chosen for this purpose the esterifi- 

 cation of asymmetric dibasic acids and the saponification 

 of their esters. He finds there are at least two different 

 of the substituting groups which act to determine 

 their influence on reaction velocities. One of these : s the 

 influence on the electrolytic dissociation : the other is 

 [firmed steric hindrance, though it is a function of several 

 single properties. Laws for the esterification of dibasic 

 larboxylic acids were based on these considerations and 

 their behaviour in actual practice exemplified : though on 

 the whole satisfactory, there are numerous exceptions, but 

 this was to be expected, as the assumption, that one 

 propertv only of the substituent influenced the reaction 

 velocitv, ..in only be a rough approximation, 



\ report on the present position of electric steel making 

 was presented by Prof. A. McWilliam. This is in type, and 

 ran be obtained at the British Association offices. The 

 report shows the progress in the electric steel melting 

 industry made during the year, whilst the actual state of 

 the industry can be judged from two tables, in which are 

 shown the furnaces, capacities, and kind of work done by 



NO. 2l8;, VOL. 87] 



the firms of Kjellin. Rochling-Rodenhauser, and Heroult. 

 In this country we have Heroult's furnaces of a united size 

 of 25 tons. A list is given of the applications of the 

 electric furnace. The report is mainly a statement of 

 industrial advance, and is somewhat disappointing from the 

 point of view of the chemist. 



The commitee appointed at the Sheffield meeting with a 

 grant to studv the influence of carbon and other elements 

 on the corrosion of steel, reports on the behaviour of a 

 series of six pure iron-carbon alloys prepared by the coke 

 crucible process at Sheffield University. Carbon exerts two 

 tvpes of influence on the corrodibility dependent upon the 

 condition of the carbide in the steel. In the rolled and 

 annealed specimens the corrodibility rises to a maximum at 

 turation point (089 per cent, of carbon), and then 

 jes upon the appearance of cementite in the steel. 

 In the hardened and tempered specimens the corrodibility 

 rises continuously from o-I per cent, to o-qii per cent, of 

 carbon, no maximum being observed at the saturation 

 point. It is cons : jered that the finer the state of division 

 of the carbide in the pearlite the greater the liability to 

 corrosion when immersed in sea water. 



The treatment previously undergone by the steel also 

 influences the solubility in acid solution. Curves are given 

 for steels treated in different ways, showing the effect of 

 ising proportions of carbon. It is established that the 

 ance offered by carbon steels when immersed in solu- 

 tions varies considerably, acording as to whether the 

 solution is of the sea-water type or is acid in character. 

 Each case must therefore be considered separately, and it is 

 impossible to specify any particulai composition or treat- 

 ment offering the best resistance to attack under all 

 conditions. 



Friday, September 1, was devoted to papers on indicators 

 and colour. Mr. Tizard's paper on the sensitiveness of 

 indicators will be published in full, and copies will be 

 available at the offices of the British Association. His 

 main conclusions are as follows : — An indicator is now re- 

 garded as a pseudo-acid or base ; the undissoi iated molecule 

 consists of two, or more than two, tautomeric forms in 

 equilibrium. This conception does not affect Ostwald's 

 method of treatment, provided that the tautomeric changes 

 that may take place are practically instantaneous. From 

 a phvsico-chemical point of view the ions may still be 

 regarded as differently coloured from the " undis- 

 sociated molecule," if it be understood by this expression 

 not one particular molecular species, but the equilibrium 

 mixture of the various forms which the indicator can 

 assume in its undissociated form. Applying Ostwald's 

 dilution law to the special case of indicators, it is easy to 

 hat when the colour of an indicator in solution is 



ai '\ midway between the extreme colours of its dis- 

 sociated and undissociated forms, then the concentration of 

 hydrogen ions (C„) in the solution must be numerically 

 equal to the dissociation constant of the indicator. Further, 

 : f this lie denoted by Ka. the colour change takes place 

 mainly between concentrations of hvdrogen ions of ioKa 

 and V1R1. It follows that the most useful indicators are 

 those which have dissociation constants not very far 

 ed from 10- 7 (the concentration of hydrogen ions at 

 the true " neutral point '"), between 10- 5 and to- 9 , for 

 example. An indicator which is a very weak acid or ba^e 

 is of no more value than one which is very strong. It is 

 of importance to know the dissociation constants of 

 indicators accurately. 



If the range of sensitiveness of an indicator is known, it 

 is possible to deduce the " end-point " of an indicator: that 

 ..iv, the probable concentration of hydrogen (hydroxy]) 

 ions in a titrating solution at the point where titration is 

 usually stopped. The end points of indicators can also be 

 found by direct experiment; as a rule it is. of course, only 

 possible" to stop a titration between certain concentrations 

 of hvdrogen ions, the extent of the range depending on a 

 number of factors. 



The accuracy of a titration of any acid by anv base 

 depends verv largely on the proper choice of an indicator. 

 This may be seen most clearly by drawing curves showing 

 the concentration of hvdrogen ions in a solution of a salt 

 when small quantities of acid or base are added in excess. 

 Erom such curves it can be deduced that an indicator which 

 has a sharp end point in anv particular volumetric opera- 



