438 



97650. 950*95. 937'S5. 93024, and 925-46 respect- 

 ively; oP,-2D, 3D, . . . 7/; arc 1025-8^, 97176, 

 94873, 936'62, 929*59. «"'! 92492. The wave- 

 lengths of oP„ - 15 are 1355-6 and 1358-6. while the 

 oP, - IS line is definitely absent. 



Sulphur. — In extending the spectrum of sulphur 

 into the extreme ultra-violet, I used dry sulphur 

 dioxide at various prtssures in both the receiver and 

 connected discharge-tube of the vacuum grating 

 spectrograph. The following is a brief summary of 

 some pf the results ol)tained. Sulphur dioxide gas 

 has a strong absorption band extending from \250o to 

 Xijoo where a narrow and relatively transparent region 

 occurs, and then another absorption band extends 

 from X1650 indefinitely into the ultra-violet. The 

 fine structure of these bands is now being studied 

 in this laboratory. By using low pressures strong 

 spectra have been obtamed even in these regions of 

 aosorption. Thus I have photographed the spark 

 and arc spectra of sulphur. The spark spectrum 

 consists of many lines and groups of lines and extends 

 to X350. The most prominent feature of the arc 

 spectrum is a number of triplets of wide separation 

 and constant frequency difference. I have classified 

 these triplets in series by analogy with oxygen. The 

 series' designation and wave-lengths are : 0P123 - iS, 

 1826-35. 1820-53. 1807-42; oP,j3-25. 1436-92. 

 1433-27. i425*ii ; 0P123-3S, 1326-69. 1323-58. 

 1316-63. oPi23-2D. 1485-53, 1481-66, 1472-99; 

 oPu3-3D, 1412-92, 1409-41, 1401-55; oPi23-4D. 

 1313-22. 1310-26 . . . (the last line of this triplet 

 was not observed as it is probably hidden by the 

 strong oxygen triplet in this region). 0P23 - 15 are 

 at 1914-96, 1900-47. and the oPj - is line is definitely 

 missing as in oxygen. The common head oPjog of 

 these series cannot be accurately calculated from' the 

 data using the Rydberg law. Fortunately the term 

 IS has been observed by Meissner [Annalen der Physik, 

 50, 713, 1916) in his study of the infra-red, hence the 

 head of these series and all the other terms may be 

 obtained at once from the above corresponding wave- 

 lengths. Thus 0P123 has the values in frequency 

 units of 82982. 83156. and 83554. Using these 

 values and the observed wave-lengths. 15=28227. 

 2S =13384. etc. Other spectra in sulphur containing 

 these terms would exist in the region of the infra-red 

 and have not yet been observed. 



Both the oxygen and sulphur spectra described 

 above show similar characteristics, namely, the in- 

 tensities of the lines and their separations are inverted 

 as compared with the known spectra of these elements 

 in the visible and infra-red, thus in the new spectra 

 the shortest wave-length of each triplet is most 

 intense, and, as indicated by the data above, the 

 frequency separation is greater between the two more 

 refrangible lines of each triplet. In both elements 

 one triplet has a missing line correspondingly placed. 

 This fact seems to indicate an inner quantum relation 

 which makes its occurrence impossible. 



On the new Bohr theory the valence level of oxygen 

 and sulphur is a Zj, hence a p or P level and my data 

 indicate this to be a triple level. Apparently there 

 exist one stable and two metastable forms of each 

 of the elements atomic oxygen and atomic sulphur. 

 On the assumption that in both elements the oPjos 

 level is the valence level, the resonance and ionising 

 potentials of the stable forms as calculated from the 

 data above are : for the oxygen atom, resonance 

 9-11 >'olts. ionisation 13-56 volts ; for atomic sulphur, 

 resonance 6-50 volts, ionisation 10-31 volts. 



J. J. HOPFIELD. 



University of California, Berkeley, 

 August 3. 



NO. 2812, VOL. I 12} 



NATURE [Septembkr 22, 1923 



[It should l>e ti> t in the foregoing coi 



munication the caj 1 rs P, 5, D refer to t 



new triplet series, and m ihc case of oxygen <■> 

 connected with the previonslv known " sii 

 series for which similar < ijs have been 



elsewhere adopted. ALso. ; terms is are 



those associated with the previously known triplet 

 series of oxygen and sulphur. — Kd. Nature.] 



Continental Drift and the Stressing of Africa. 



Mr. Wavland in Nature of August 25, p. 279, 

 brings forward weighty arguments, based <m the 

 results of the Geological Survey of Uganda, to rebut 

 the usually accepted view that the Protectorate, 

 like most of eastern Africa, and probably western 

 .•\frica as well, has been predominantly in a state of 

 tension. I shall be surprised, however, if further 

 work does not disclose the existence of at le 



normal faulting with a north and south strik ig 



the former existence of east and west teusiun. It 

 may well be that compression and tension have more 

 than once alternated with each other in Uganda. 

 There is no reason, too, why a change of conditions 

 may not convert a true rift valley formed in a period 

 of tension into one bounded by reversed faults. 



I am by no means prepared to admit that the birth 

 of the moon (supposmg it to have in fact arisen by 

 separation from the earth) must have necessarily 

 been a " piece of extremely ancient history." Sir 

 George Darwin gives reasons for his belief that it 

 took place considerably more than 50 or 60 million 

 years ago. Now Dr. Holmes's calculations, based on 

 the uranium-lead ratio of certain minerals, show an 

 antiquity of about 500 million years for the beginning 

 of the Cambrian. This would suggest that the 

 interesting event in question may have occurred at 

 some time within the limits of the fossiliferous recor 

 Nor is there any reason to believe that it mu 

 have been marked by stupendous catastrophic dis- 

 turbances. A sphere of the earth's size yields itself 

 slowly but practically unresistingly to a force acting 

 continuously upon it — in this case the centrifugcU 

 force due to its rotation, accelerated more \>\- 

 the progressive condensation of its interior tha 

 retarded by the tidal action of the sun. 



The process of separation may have been protracted 

 over a considerable time, more perhaps than that 

 represented by a single geological formation. Indeed, 

 there is a great deal to be said for the suggestion that 

 it may have commenced about the middle of th 

 Carboniferous and continued till the close of tl. 

 Trias. This would account for the fact that in tl; 

 portion of the earth's crust which has been chieli\ 

 studied, that is to say the extra Pacific area, there 

 appears to have been throughout that lengthy period 

 a general, though by no means a complete, recession 

 of the ocean, which would presumably be attracted 

 towards the protruding mass of the moon. At some 

 stages of this emergence the bulk of the atmosphere 

 would be affected in the same way, giving rise in 

 the region antipodal to the moon to a period of 

 marked rarefaction and cold, resulting in the Talchir 

 and Dwyka ice-age which has been recognised not 

 only in South Africa, but also in India, Australia 

 the Falklands. and South America, all formerb. 

 according to Wegener, and I believe he is right, 

 clustered about Africa. If there be any truth in 

 this supposition we should expect the chief period 

 of tension in Africa and its surroundings to have 

 existed in Mesozoic and early Kainozoic times, not 

 in the Palaeozoic or pre-Cambrian. The powerful 

 tidal action of the moon, while still comparatively 



i 



