772 



SCIENCE 



[N. S. Vol. XL. No. 1039 



C. T. Knipp: Experimental Data on the Stability 



of Positive and Negative Ions. 



In an investigation by Dr. O. H. Smith and the 

 writer of the properties of the retrograde rays 

 from a Wehnelt or hot lime cathode, it was found 

 necessary in order to make their presence known, 

 to accelerate these rays by passage through a 

 strong electrostatic field. The photographic 

 method of J. J. Thomsoni was employed. 



A number of plates show anomalies as to the 

 direction of deflection, i. e., the appearance of 

 positive lines with negative acceleration, while the 

 same exposure shows but little or no trace of the 

 negative lines. This is contrary to what is ex- 

 pected under the given conditions. 



This anomaly, the appearance of positive lines 

 on the photographic plate when the acceleration is 

 such as to allow only negative ions to get through 

 the accelerator into the electric and magnetic de- 

 flecting fields beyond, can be satisfactorily ex- 

 plained if it is assumed that the positive ion is 

 more stable than the negative ion. In other words 

 a negative ion loses an electron more easily than 

 does a positive ion. 



The paper in detail follows the path of the 

 negatively accelerated ion as it issues from the 

 accelerator and notes the possible changes that 

 evidently take place (upon the supposition stated 

 above) as it moves on through the deflecting fields 

 to the photographic plate. In this way every part 

 of the line on the photograph (i. e., the straight 

 portion as well as the parabolic portion), is satis- 

 factorily accounted for, and hence the conclusion 

 that the positively charged ion is more stable than 

 the negatively charged ion. 



A number of photographs showing these lines ac- 

 company the paper. 



III. Physics 

 A. A. MicHELSON: Behavior of Metals and Other 

 Substances Near the Rupture Point. 



E. A. MiLLiKAN: The Coefficient of Slip in Gases 

 and its Belation to the Nature of the Impact 

 between a Molecule of a Gas and the Surface 

 of a Solid or Liquid. 



In 1911 I brought forward a new methods for 

 the very accurate evaluation of the coeflicient of 

 slip between a gas and the surface of a liquid or 

 solid. This coefficient was shown to be equal to 

 the quantity Al in the equation for the law of fall 

 of a small sphere through a gas.2 This quantity 



1 J. J. Thomson, ' ' Bays of Positive Electricity, ' ' 

 1913. 



2 Physical Eeview, XXXII., p. 382, 1911. 



Al was in turn shown to be proportional to the 

 slope of the line obtained by plotting e f 

 against l/pa in the ' ' droplet ' ' method for the 

 determination of e. The values of this slope have 

 now been obtained with different gases and dif- 

 ferent kinds of droplets. It has hitherto been 

 supposed from the work of Kundt and Warburg 

 that the coefficient in question is in all cases 

 proportional to the mean free path of the gas 

 molecule. This conclusion is now shown to be 

 incorrect; for the above-mentioned slopes are not 

 only found to depend on the nature of the sur- 

 face against which the gas molecule impinges 

 when this molecule remains the same, but also 

 upon the value of the impinging molecule when 

 the surface is the same. These results show that 

 in general gas molecules are not ' ' diffusely ' ' re- 

 flected from liquid and solid surfaces as they 

 have recently been assumed to be by Knudsen 

 and others.3 



IV. Chemistry 

 C. W. Balke and Geo. W. Sbaes: The Atomic 



Weight of Tantalum. 



In order to determine this constant two ratios 

 were studied, 2Ta,Gl5: Ta^O^ and TaClj: 5Ag. 



In the study of the ratio 2TaClB: Ta^O», the 

 tantalum chloride was hydrolyzed, using nitric 

 acid, and the resulting tantalum oxide was evapor- 

 ated to dryness and ignited. In five determina- 

 tions, however, a constant weight was not obtained 

 even after repeated ignition. A fine white deposit 

 was found in the exit tubes of the reaction flask 

 after each analysis showing that the oxide was 

 being lost. An examination into the cause of 

 this showed that the hydrochloric acid was com- 

 pletely removed by ignition, that the nitric acid 

 was difficultly removed, if at all, and that tanta- 

 lum oxide was lost either mechanically or by 

 volatilization, all of which indicated that the 

 method was unsatisfactory. 



In the study of the ratio TaClB: 5Ag, several 

 methods involving the removal of the hydro- 

 chloric acid from the tantalum were tried but 

 without success. The only method found satis- 

 factory was to dissolve the tantalum chloride in 

 an approximately 5.5 per cent, hydrofiuoric acid 

 solution and to precipitate the resulting hydro- 

 chloric acid in the presence of the dissolved 

 tantalum. Platinum vessels were used through- 

 out and the final end point was determined in the 

 nephelometer. 



3 Annalen de PhysiTc, Kundsen Papers, 1909- 

 1912. 



