August ig, 1922] 



NA TURE 



249 



25846 25849 

 26053 26047 



The ^-region beyond 26047 has thus far not been 

 swept systematically. When this is done, I have but 

 little doubt that ninety or more of the one hundred 

 and five lines of helium will be accounted for. In 

 these circumstances one would feel justified in assert- 

 ing that the absence of mutual repulsion between the 

 electrons is not (as I first thought) an exception but 

 rather the rule. A simple estimate will show that if 

 the usual Coulomb repulsion law were valid in any 

 of the considered stationary states, the mutual energy 

 of the electrons would contribute several thousand 

 units to v. Since it is hard to explain away so many 

 coincidences as due to chance, we are driven to the 

 belief that the electrons within the atom do not repel 

 each other even with a small fraction of the force 

 usually attributed to them. In other words, the 

 field of force of a bound electron seems to be entirely 

 engaged by the nucleus, at least in the case of helium 

 and probably of lithium, but possibly also in that of 

 the higher atoms. Ludwik Silberstein. 



129 Seneca Parkway, Rochester, New York, 

 July 26. 



The Primitive Crust of the Earth. 



In reference to the letter of Dr. Harold Jeffreys 

 (Nature, July 29, 1922), I wish at once to say that 

 nothing in my letter published on July 8 was intended 

 to express my adhesion or non-adhesion to those who 

 support the planetesimal hypothesis. Even if we 

 think that the earth originated in a rain and con- 

 centration of solid planetesimals, we may, with Prof. 

 R. A. Daly, regard its complete fusion at a later 

 stage as a very probable event. At some time or 

 other, the earth may well have possessed a crust 

 consolidated from " igneous " fusion. Prof. J. Joly 

 now suggests to us, with his unfailing brilliance of 

 outlook, the recurrence of such a crust after successive 

 meltings of the globe. What I have urged, however, 

 is that the oldest rocks traceable by geologists must 

 not be regarded as a record of a primitive crust. 

 They are sediments, invaded again and again by 

 igneous matter from below. We cannot conclude 

 from our Archaean schists, which are so often con- 

 verted into composite gneiss, that there was ever a 

 crust formed of crystalline rocks about the globe. 

 The " extent of the crust accessible to geologists " 

 is, of course, much more than the film 2-5 km. thick 

 stated by Dr. Harold Jeffreys. Owing to the great 

 movements that bring up antique masses from the 

 depths, rocks that consolidated finally under several 

 miles of sediments now form a large part of the 

 surface. But so far no planetesimal sediment has 

 come to light, although matter of the mineral com- 

 position demanded by the hypothesis is associated 

 with many igneous upwellings. 



NO, 2755, VOL. I IO] 



In support of the concluding remarks of Dr. Harold 

 Jeffreys, attention may be directed to "A Critical 

 Review of Chambertin's Groundwork for the Study of 

 Megadiastrophism," by W. F. Jones, published in 

 the American Journal of Science for June of the 

 present year. Grenville A. J. Cole. 



Carrickmines, Co. Dublin, July 29, 1922. 



Peculiarities of the Electric Discharge in Oxygen. 



Several years ago I described (Phil. Mag., April 

 1908) a discontinuity in the electric discharge in 

 oxygen at pressures near to o-8 mm. Namely, when 

 a current (0-0025 amp.) was passed in a discharge 

 tube (diam. 2-4 cm.), the electric force in the positive 

 column suddenly changed on slightly lowering the 

 pressure from about 11 volts per cm. to about 20, 

 an effect which could be reversed by raising the 

 pressure. 



Some experiments which I have made recently, 

 with the assistance of Mr. E. P. Cardew, have shown 

 that at pressures in the same neighbourhood, with a 

 fixed circuit (battery and resistance), the discharge 

 is not uniquely determined, but can be one or the 

 other of two distinct types, distinguished by a re- 

 markable difference in the values of the electric force 

 within the positive column, one of these values being 

 about twice the other. The magnitude of the current 

 with the higher electric force in its positive column is 

 less than the other, since the potential difference of 

 the electrodes is greater in its case ; but the currents 

 tend to equality when the electrodes are so near 

 that the positive column tends to disappear. The 

 two discharges differ only slightly in appearance : 

 the positive column of the smaller one with the 

 higher electric force being somewhat shorter and a 

 trifle paler than the other — both being without striae 

 in general. 



To give an example. With a battery of 990 volts 

 and external resistance 363,000 ohms, electrodes 21-8 

 cm. apart in a tube of 27 mm. diameter, and pressure 

 0-75 mm., the currents observed were 1-19 and 0-883 

 milliamperes ; their positive columns were nearly 

 15 and 14 cm. long, and the electric force within 

 them about 9 and 18-5 volts per cm. respectively. 

 In this and in many other cases, by means of a certain 

 arrangement, the discharge could be made to change 

 from one form to the other without stopping the 

 current or altering the circuit. 



The region of pressures within which alternative 

 currents have been so far observed are from 0-64 to 

 0-91 mm. The two types of discharge differ in 

 stability according to the pressure and the magnitude 

 of the currents, so that the discharge tends to assume 

 one type rather than the other. But the one having 

 the high electric force in the positive column is much 

 more definite and invariable than the other for a 

 given pressure, being in this respect similar to dis- 

 charges in other gases, so that that electric force can 

 be determined with much more precision, and is in 

 fact nearly the same as in hydrogen. 



Since these effects hold good for a large range of 

 current, it is obviously possible, by adjusting the 

 external circuit, to make two discharges of the same 

 arbitrary magnitude (of the necessary order) pass 

 through oxygen, between electrodes at a given 

 distance apart, at any given pressure within the 

 range in question, one of which will have the high 

 electric force in its positive column and the other 

 the low. P. J. Kirkby. 



Saham Tonev Rectory, Watton, Norfolk, 

 July "26. 



