512 



SCIENCE 



[Vol. LVI, No. 145.3 



Dr. LjTiian will enter upon his new duties on 

 January 1. 



Propessoe Leroy Patton, foi-merly of 

 Muskingum College, Ohio, lias been appoin1)ed 

 associate geologist of the Bureau of Eeonomic 

 Geology in the University of Texas. Dr. E. H. 

 Sellards, who has been with ithis bureau several 

 years, has been promoted to be chief geologist, 

 Professor T. L. Bailey, from the University of 

 California, has aeeeipted tJie position of as- 

 sistant geologist, vacated by Professor W. S. 

 Adkins a year ago, and Miss Dorothy Shoaf, 

 from the University of Chicago, has been ap- 

 pointed cm-ator of the collections. 



Db. J. L. Shellsheae, of Sydnej', AustraJia, 

 has been appointed to the chair of anatomy in 

 (the new College of Medicine of Hongkong Uni- 

 versity. 



DISCUSSION AND CORRESPOND- 

 ENCE 

 NOTE ON THE DISSOCIATION OF CARBON 

 IN THE INTENSIVE ARC 



Some two years ago while experimenting 

 with the extremely powerful arcs used in the 

 Sperry search lights we noted the singular 

 color and peculiar spectrum in the "negative 

 tongue" which appears at currents of 100 am- 

 peres and upwards. It develops rather sud- 

 denly as a core of the negative flame, suggest- 

 ing the inner cone of a blast lamp save in 

 color, which is pale purplish. 



The spectroscope disclosed a small number 

 of clearly marked lines superimposed on 

 fainter hazy and complex bands, due to the 

 surrounding arc flame. Examining the tongue 

 spectrum of the lines from time to time, we 

 found substantially the same sj)eetrum from 

 various makes of unmineralized carbons, for- 

 eign and domestic. Finally, using a five inch 

 afihromaiie condenser to throw the image of 

 the tongue on a ground g'lass we examined it in 

 detail with a direct vision spectroscope 

 equipped with a scale, comparison prism and 

 holder for spectrum tubes. 



We thus found as characteristic of the 

 tongue spectrum, some fifteen well defined lines. 

 Of these, seven were good coincidences witli 

 the most conspicuous of the well known helium 

 lines, and two others wilii Ha and H/J. 



The He lines were wl: 7066, 6678, 5876, 

 5048, 5016, 4922, 4388. 



Five of these lines belong to the single line, 

 so-called Parhelium series, being the brightest 

 lines of the principal and second subordinate 

 series, and tie three of the first subordinate 

 series. 



Of the doublet system ;the chief lines of the 

 sharp and diffuse series respectively appear, 

 not very brilliantly, while we have not yet 

 detected any of the enihanced series. All indi- 

 cations point to the dissociation of a certain 

 proportion of carbon nuclei with the conse- 

 quent appearance of He due ito the immense 

 concentration of energy in this 150 ampere 

 arc. The H lines may (be due to water-vapor 

 absorbed by the soft carbon core, or perhaps 

 to further dissociation. We are now setting 

 up a concave grating spectrograph for the 

 closer examination of the tongue spectrum 

 under much higher dispersion, and hence with 

 a less obtrusive background. We hope that the 

 evidently very high ionization power here 

 manifested may lead us to interesting develop- 

 ments with still other elements. 



Our thanks are due to the courtesy of Mr. 

 Sperry in extending the great facilities of his 

 laboratory. 



Louis Bell 

 P. R. Bassett 



THE DETERMINATION OF FA.T IN CREAM 

 To THE Editor op Science: The authors 

 (E. G. Mahin and E. H. Carr) of a paper on 

 "Errors in the Determination of Fat in 

 Cream," read at the Birmingham meeting of the 

 American Chemical Society, have experienced 

 considerable surprise at the tone of a letter by 

 H. W. Gregory, appearing in the issue of 

 Science for September 15, 1922, in which he 

 discusses our work upon this subject. Profes- 

 sor Gregory has based his criticisms upon a 

 mere advance abstract, containing no details 

 of experiment or reasoning, and without ade- 

 quate knowledge of the real points at issue. 



In the original paper by Mahin and Carr 

 (not yet published) we have simply called at- 

 tention to a hitherto unsuspected error in the 

 almost universally used "glymol" method for 

 making fat readings in the Babcoek tests on 



