746 



NATURE 



[June 



o, 1922 



Dynamics," § 46). For several electrons there is no 

 similar simple transformation. The " radiating " 

 electron compels the remaining electrons to execute 

 motions of 'reaction, which also influence the 

 nucleus. 



Probably it will be possible to derive a sufficiently 

 approximate formula for the case of the p- and d- 

 motions of lithium. This must be very difficult, 

 however, for the case of the 1-5 S-path. The fact 

 that at all events equation (i) cannot be true in general 

 for atoms with several electrons will be shown by 

 the following example (though of course on account 

 of the Principle of Correspondence not representing 

 a process really occurring in nature) : two electrons 

 move around the nucleus in a centrally symmetrical 

 configuration, at first in a two-quantum and then in 

 a one-quantum circle. By symmetry the nucleus 

 remains continually at rest. In this case therefore 

 the liberated energy and, consequently, the radiated 

 V would, contrary to (i), be exactly independent of 

 the mass of the nucleus. P, Ehrenfest. 



The University, 



Leyden, Holland. 



I SHOULD like to add a few remarks to the interesting 

 letter of Prof. Ehrenfest about the contents of which he 

 was so kind to inform me before publication. As 

 pointed out in his letter, the effect of the mass of the 

 nucleus on the spectrum of an atom, containing more 

 than one electron, is a complex problem which depends 

 on the electronic arrangement in the states of the 

 atom, involved in the emission of the lines, in a way 

 which has hitherto not received sufficient attention. 

 Not only may the mass effect disappear completely 

 in such cases, where several electrons move round the 

 nucleus in equivalent orbits, but, as indicated by 

 Prof. Ehrenfest, this effect may also in case of the 

 motions which we actually meet in the emission of the 

 series spectra be different from that calculated for an 

 atom with one electron. 



Although in the emission of these spectra we are 

 concerned with motions whereby a single electron 

 moves in an orbit different in type from the orbits 

 of the other electrons, the problem differs essentially 

 from the problem of two bodies in celestial mechanics. 

 Thus according to the picture of atomic constitution, 

 outlined by the writer in two letters to Nature 

 (March 24, 1921, October 13, 192 1), we shall assume 

 that the electron connected with the emission of the 

 series spectra, although during the larger part of the 

 revolution it remains outside the configuration of the 

 electrons in inner groups, it will nevertheless in certain 

 states penetrate into the interior of the atom during its 

 revolution. The fact that the electron in the inner 

 loop of its orbit is subject to large forces is of pre- 

 ponderant influence as regards the fixation of the 

 energy in the corresponding stationary states of the 

 atom. For such a motion the effect of the nuclear 

 mass might differ essentially from that estimated 

 from an examination of the mechanical properties of 

 the motion in the outer loop only, and the question 

 arises, whether the mass effect is sufficiently large to 

 account for the discrepancies, observed by Merton, 

 in the wave-lengths of certain lines in the spectra of 

 lead isotopes, which although very small are yet much 

 larger than those to be expected from the simple 

 formula quoted in Prof. Ehrenfest's letter. 



Although this question seems difficult to settle 

 without a closer investigation, it would scarcely 

 appear probable that the answer will be affirmative. 

 On the other hand, it cannot be excluded that the 

 discrepancies in question are due to a slight difference 

 in the field of force surrounding the nucleus, arising 

 from the difference in the internal nuclear structure 



of the lead isotopes. This possibility has been dis- 

 ciissed from various sides. At first sight we meet 

 with the difficulty, that the dimensions of the nucleus 

 (ca. 3 X 10-^2 cm.), estimated from experiments on the 

 scattering of a -particles, are exceedingly small in 

 comparison with the dimensions of the orbits of the 

 electron responsible for the emission of the series 

 spectra, which are of the order lo"* cm. or larger. 



This difficulty may disappear, however, by consider- 

 ing the circumstance mentioned above, that in certain 

 states the series electron during a short interval of its 

 revolution penetrates deeply into the interior of the 

 atom. In fact, we must assume, that this electron 

 in the states corresponding to the S-terms of the 

 series spectra penetrates to even smaller distances 

 from the nucleus than the electrons in the innermost 

 group of the atom, the dimensions of which are 

 in lead smaller than lo"" cm. To the possible im- 

 portance of this point in connection with the spectra 

 of isotopes my attention was kindly directed by 

 Dr. Kramers in a discussion about Prof. Ehrenfest's 

 letter. N. Bohr. 



University, Copenhagen. 



The Destruction of Mosquito Larvae 

 in Salt or Brackish Water. 



A NUMBER of experiments on the destruction of 

 mosquito larvae by the well-known system of " surface 

 oiUng," carried out at Hayling Island during the 

 year 1921, supplied further evidence of the fact that 

 this method is not one of universal application. 

 The production of an unbroken film of a sufficiently 

 lasting nature is sometimes an impossible task, 

 notably in cases where the water surface is broken 

 up by growths of reeds, etc., or is too freely exposed 

 to the wind. In cases of this kind it is necessary to 

 discard the oil film in favour of a " larvicide," that 

 is to say, a substance which, by mixing with the 

 infested water, will destroy the larvae. 



Unfortunately, however, practical information 

 concerning larvicides is difficult to obtain. In the 

 literature of the subject references are to be found 

 to a number of suggested substances, but the vast 

 majority of these stand self -condemned owing to the 

 prohibitive cost that would be entailed by employing 

 them in the prescribed " strengths " on any practical 

 scale. A large number of these larvicides are, 

 moreover, admittedly ineffective when added to 

 brackish or salt water, and are consequently of 

 little value in districts such as Hayling, where the 

 salt-water mosquito, Ochlerotatiis detritus, is the 

 principal offender. 



A number of tests have recently been carried 

 out in the laboratory of the Hayling Mosquito Con- 

 trol, in the hope of discovering a larvicide which 

 could be used successfully (and at a low cost) in 

 salt, or partly salt, water. It was found that a 

 Hquid containing 15 per cent of soluble cresol, sold 

 as a disinfectant under various names, gave very 

 promising results in the laboratory. This liquid, at 

 a dilution of i in 16,000, was found to kill the larvae 

 of Ochlerotatiis detritus in one hour ; at a dilution 

 of I in 32,000, in one and a half hours ; and at a 

 dilution of i in 48,000, in three and a half hours. 

 In the majority of these experiments the water 

 containing the larvae was of a salinity about half that 

 of sea water. 



In order to test this larvicide on a larger scale, a 

 shallow stretch of brackish water adjoining the 

 Hayling Golf Links was selected for experiment. 

 This water was very heavily infested with the larvae 



NO. 2745, VOL. 109] 



