December 8. 1923] 



NATURE 



September 1923, 37,000 tons were manufactured by 

 the acid, and 9000 tons by the basic Bessemer process. 

 Very large quantities of basic Bessemer steel are being 

 made in Germany and Belgium at the present time. 



So far, therefore, as the above quotation relates to 

 the Bessemer process, it is entirely inaccurate and the 

 revisers are quite justified in giving details. The 

 funeral of the Bessemer process has frequently been 

 predicted, but it has never taken place. 



H. C. H. Carpenter. 

 Royal School of Mines, 

 South Kensington. London, S.W.7, 

 November 19. 



Prof. Carpenter is evidently right, and I am glad 



lat he has corrected my mistake in reference to 



e Bessemer converter, — the statement as to the 



blanc process was, I believe, correct. It would be 



interest, however, if Prof. Carpenter could give the 



ite of construction of the last new Bessemer plant 



ected in Great Britain for steel manufacture. If 



\v plants are not being constructed, the view that 



le Bessemer process for steel is really "of historical 



lerest only," would not be altogether unjustified, 



.^iuce this process would then rank, like the hansom 



cab, as one of the products of the Victorian age, of 



which the usefulness is likely to diminish rather than 



to increase in the twentieth century. 



The Reviewer. 



The Spectra of Fifth Group Metals. 



We have photographed the absorption spectrum of 

 bismuth and also the spectrum of the thermionic 

 discharge at potentials ranging between 4 and 60 

 volts. Several stages in the excitation of the arc 

 spectrum, and at least two classes of spark lines, have 

 been recognised ; 64 arc lines have been classified. 

 The spectrum of the neutral atom is characterised by 

 wide doublets, and most of the energy-levels so far 

 identified are of /j-type. 



Electrical measurements of the arcing potential and 

 potentials of inelastic impact were made by two of 

 the authors and the late Dr. Oswald Rognley in 1918. 

 They found inelastic collisions at intervals of 2-0 ± 0-2 

 volts and ionisation at 8-o ± 0-5 volts. The interpreta- 

 tion is as follows : 



The first resonance potential, 2-0 volts, represents 

 the mean of the excitation voltages for several weak 

 spectral lines of the type mp - np'. At 4-0 volts, we 

 obtain the strong raies ultimes, XX 3067 and 4722 A.U. 

 Excitation stages above 5 volts are difficult to sepa- 

 rate. The first spark spectrum appears near 14 volts. 



The absorption spectrum at 8oo°-iooo° C. shows 

 lines due to the atom, and prominent bands which 

 have not been described previously. A group of 

 seventeen bands lies between 2874 and 2672 A.U., 

 while a second group extends from 2205 A.U. toward 

 shorter wave lengths. At lower temperatures the 

 bands disappear though the lines still may be re- 

 cognised. Ihey lie at 3067, 2276, 2230, 2228, and 

 1954 A.U., and all originate on the lowest energy- 

 level of the atom. No absorption lines arising from 

 other levels were observed, even at a temperature of 

 1050° C. 



Practically all the arc lines of arsenic between 3119 

 and 2000 A.U. can be classified by means of constant 

 differences found by Kayser and Runge (Ann. d. 

 Physik, V. 52, 1894). We have discovered a few 

 additional classifications. This spectrum is remarkable 

 in that it possesses no lines in the visible region. There 

 is a range of 38,000 cm.-* and another of 32,000 cm.-* 

 in which no energy levels have been found. If there 

 are energy levels in these regions, they can probably 



NO. 2823, VOL. 112] 



be detected only by the discovery of new lines, or 

 the utilisation of lines at present listed in the spark 

 spectrum. The potential of inelastic impact, 47 volts, 

 given by Foote, Rognley, and Mohler {Phys. Rev., 

 13. 59. 1919) corresponds to the mean of the wave 

 numbers of the raies ultimes. The classification of 

 the spectrum shows that the ionisation potential must 

 be at least io-6 volts, while the experimental value 

 is 11-5 volts. 



Arthur E. Ruark. 



F. L. Mohler. 



Paul D. Foote. 



R. L. Chenault. 

 Bureau of Standards, Washington, D.C., 

 November 8. 



Tracts for Computers. 



I REGRET that certain errata have been found in 

 No. III. of the above Tracts. As they might cause 

 confusion to any one computing from one of the 

 formulae affected, I have had an erratum slip printed, 

 which can be obtained by purchasers of the above 

 series by sending a stamped and addressed envelope 

 either to Mr. C. F. Clay, Cambridge University Press, 

 Fetter Lane, E.C.4, or to The Secretary, Biometric 

 Laboratory, University College, Gower St., W.C.i. 



Karl Pearson. 

 Biometric Laboratory, 

 University College, London, 

 November 17. 



Mesozoic Insects of Queensland. 



Lest the reference in Nature of July 7, p. 20, to 

 Queensland Geological Survey Publication, No. 273, 

 ma}'- lead readers to think that the account of the 

 Coleoptera is the first published work on the insects 

 from the six-inch seam containing insect remains at 

 Ipswich, I would point out that a series of papers 

 dealing with these insects has already been published 

 by Dr. R. J. Tillyard (Queensland Geol. Survey, Pub. 

 253, 1916 ; and " Mesozoic Insects of Queensland," 

 Nos. I to 9, Proc. Linn. Soc. N.S.W., 1917 to 1922). 



A. B. Walkom, 

 Secretary. 

 Linnean Society of New South Wales, 

 Sydney, October 2. 



[The paragraph to which Dr. Walkom refers was 

 intended to direct attention to a particular piece of 

 work, and no attempt was made to mention earlier 

 publications on the same subject, though the con- 

 tributor was familiar with them. — Editor, Nature.] 



Hafnium or Jargonium. 



The recent discovery of hafnium in minerals con- 

 taining zirconium serves to remind us of the discovery 

 of jargonium by Sorby in 1869 (Chem. News, vol. 20). 

 He found that many zircons contained as much as 

 10 per cent, of the new element. The two closely- 

 related elements, zirconium and jargonium. could be 

 most readily distinguished by spectroscopic methods. 

 Sorby and Forbes found that there was such a marked 

 difference in the solubilities of the chlorides in strong 

 hydrochloric acid, that it was possible to make a 

 qualitative separation. Three years later Cochran 

 investigated this subject and suggested that zirconia 

 and jargonia were identical. My object in bringing 

 this matter before readers of Nature is to suggest 

 that the work of Sorby may possibly entitle him to 

 rank as the discoverer of the new element of atomic 

 number 72, and that jargonium may have priority 

 over hafnium and celtium. T. L. Walker. 



University of Toronto. 



