8i2: 



NA TURE 



[June 24, 1922 



where ^i is the frequency at the theoretical Umit of 

 the Balmer series, 



I'a is the frequency of the highest line that can 

 be formed, 



vj is the frequency at the redward edge of the 

 outlying spectrum. 

 This expression tells us that, expressed in frequencies 

 (and the relation holds approximately for wave- 

 lengths), the margin of overlap of the continuous and 

 line spectra should be equal to the interval between 

 the highest observable line of the Balmer series and 

 the theoretical limit. 



There are two factors tending to modify the above 

 conclusion. One is that, by the kinetic theory, 

 comparatively few electrons of approximately zero 

 velocity are to be expected. Since these determine 

 the redward edge of the outlying spectrum, that edge 

 should be faint, and the effective limit might be of 

 slightly less wave-length. The other is the fact that 

 all the atomic systems will not at any one instant 

 be reduced by molecular interference to outer orbits 

 of exactly the same order. In these circumstances 

 we should expect the large systems to determine 

 the highest visible lines, and the small ones to establish 

 the redward edge of the outlying continuous spectrum. 

 The effect here would be to introduce a " blurring " 

 factor, and increase the overlap. While these two 

 factors operate against each other, it seems quite 

 impracticable to attempt an estimate of their net 

 effect. 



To check the conclusions, the data on the emission 

 spectrum of the chromosphere and the absorption 

 spectra of a Cygni and Vega are collected in the 

 accompanying table. Unfortunately the emission 

 spectra of the nebute have not been measured accur- 

 ately enough to establish their limits. In our theo- 

 retical discussion we have regarded the problem from 

 the point of view of emission, and it is perhaps not 

 entirely justifiable to check the conclusions through 

 recourse to absorption spectra, for absorption and 

 emission cannot, in such a case, be regarded as 

 exactly complementary. It is necessar}^ however, 

 to use the stellar spectra, since, with that of the 

 chromosphere, they constitute the only radiations 

 that have been sufficiently well observed. 



* The most refrangible line observed by Evershed was 365iA. Mitchell, 

 on the occasion of the 1905 eclipse (Astrophys. Jour., 38, 431, 1913 : also 

 Publ. Leander McCormick Obs., 2, 49), photographed six additional lines, 

 the highest being at 3656.8A, the value adopted here. Evershed's plates 

 were taken with a prism spectrograph and Mitchell's with a grating instru- 

 ment of considerably greater power. The greater extent cf the latter's 

 spectrum is no doubt due to that fact. 



The last column contains the positions of the 

 lower edge of the outlying spectrum computed from 

 the frequencies of the highest visible lines, using 

 equation (3). The discrepancy between the observed 

 and derived values for Vega is rather large and 

 undoubtedly exceeds the error of measurement. It 

 should be remembered, however, that in the atmo- 

 sphere of a star absorption must take place throughout 

 a considerable range of density, corresponding to 

 different levels. The higher lines of the series, on 



transfer from infinity to the nth orbit. This is also equal to the kinetic 

 energy of the electron due to motion in its orbit, so that an electron at 'rest 

 at the nth orbit will have lost twice that amount. Therefore 



V, = N(l/2^-2/«2). 



Remembering that i',=N/2*, the relation (3) follows. 



which the computations rest, probably originate in 

 the upper and rarer atmosphere, where the conditions 

 are favourable for their formation, while we should 

 expect the edge of the outlying spectrum to be 

 determined in a region of comparatively high density. 

 This would account for a divergence such as the one 

 shown. Considering the numbei of extraneous factors 

 that have a bearing on the problem the agreement 

 is probably as good as might be expected. The 

 measurements show the progress of the edge of the 

 outlying spectrum toward the redward as the higher 

 members of the Balmer series fade out, and this is 

 in general accord with the theoretical deductions. 



Summary of Conclusions. — The outlying continuous 

 spectrum found in certain celestial spectra beyond 

 the limit of the Balmer hydrogen series is, as was 

 suggested by Evershed, almost certainly due to 

 hydrogen. It should be more completely studied 

 astronomically, and .serious effort should be directed 

 toward developing it in the laboratory, on account 

 of its theoretical interest and of its bearing on 

 astronomical problems. 



The spectrum is explicable, on the basis of the 

 Bohr theory, as resulting from a change of state, 

 as between a free electron and one in the second 

 orbit. Bohr's original application of the principle 

 to the case of absorption through photoelectric 

 action on hydrogen atoms, is extended to the con- 

 ception of emission as resulting from the capture of 

 free electrons by hydrogen nuclei. 



Theory and observation are in accord in placing 

 the beginning of the outlying spectrum, not at the 

 theoretical limit of the Balmer series, but to the 

 redward of it, the amount of the displacement being 

 greater as the number of observed series hues is less. 



It seems possible to account for the spectrum on 

 the basis of either thermal or electrical excitation. 

 Regarded as a heat effect, it indicates for the planetary 

 nebulae a temperature of the order of 1000° Abs. 

 Cent, or more. (The figure is not, however, offered 

 as an estimate of the temperature of the nebulee.) 



W. H. Wright. 



Mount Hamilton, California, April 21, 1922. 



Discoveries in Tropical Medicine. 



I HAVE never thrown any doubt upon the influence 

 of the suggestions made by Manson to Ross which 

 led to the close study by Ross of the carriage of the 

 malaria parasite (of both birds and man) by mos- 

 quitoes, and the discovery by him that mosquitoes 

 of the kind known as Anopheles and not those 

 of the kind known as Culex are the " intermediate 

 hosts," in which the parasites causing malaria in 

 man undergo necessary and remarkable stages of 

 their development. 



I was a member of the Committee of the Royal 

 Society with which Ross was in constant communica- 

 tion during his work in India, and followed that work 

 step by step in the reports sent home by him. My 

 knowledge of the work of Laveran, of Labbe, Dani- 

 lewski, and of Celli and of Grassi and the Italian school 

 does not support the claims to "discovery " put forward 

 on behalf of Manson by some of his friends. They 

 are exaggerated and inaccurate — though Manson's 

 influence and enthusiasm need no such mistaken 

 advocacy in order to receive recognition. 



It is the fact, in spite of assertions to the contrary, 

 that Manson did not discover the part played by 

 the mosquito in the transmission of Filaria sanguinis 

 hominis. Important details as to the part played 

 by the mosquito — of whatever kind — in that trans- 

 mission have yet to be ascertained. They are still — 

 at the present moment — a subject of investigation. 



NO. 2747, VOL. 109] 



