November 25, 1920] 



NATURE 



407 



and (C) the posterior molar, all of the left ramus of the 

 mandible. In (B) and (C) can be seen the isthmus 

 joinin}^ the roots of the molars and the traces of the 

 original third root. The peculiar cusp-like pattern on 

 the face of the teeth is also well shown. 



It should be mentioned that the knob which is 



Fio. 3. 



apparent in front of the nasal region of the skull in 

 Fig. I is merely a concretion difficult to remove with- 

 out damage to the ^ull. 



T. Thomso.v Flvnn, 

 Ralston Professor of Biology. 

 University of Tasmania, Hobart, September 9. 



The Energy of Cyclones. 



1 DO not find that people in general are aware of 

 an important source of energy for the maintenance 

 and intensification of cyclones, nor am I acquainted 

 with a clear exposition by a nieteorol^ist that the 

 condensation of aqueous vapour will suffice. 



Atmospheric pressure being a ton weight per square 

 foot, the disappearance or collapse of a cubic foot of 

 ordinary air would yield a foot-ton of work. The 

 disappearance, by complete condensation, of the 

 iiqueous vapour in 760/127, say 60, cubic feet of 

 atmosphere would yield tlie same' amount. 



If, then, the temperature of saturated air fell from 

 18° to 12° C. by reason of condensation and rainfall, 

 so that the vapour-pressure diminished from 15-36 to 

 10-46 mm. of mercury, a foot-ton would be generated 

 in each 155 cubic feet of that region of the atmo- 

 sphere. Incidentally, the corresponding deposit of 

 liquid would be j grams per cubic metre, or a rain- 

 fall of J in. from a vertical mile of air. 



.\ssuming that the above fall of temperature in the 

 central region of a travelling cyclone is not excessive, 

 the energy available In each cubic mile of it would 

 be nearly a thousand million foot-tons. 



Oliver Lodge. 



With reference to Mr. R. M. Declcy's letter on the 

 above subject in Natitrk of November 11, may I 

 suggest that the energy of a cyclone is derived from 

 the heat-energv of the earth's surface? If we assume 

 that the air which asrends In the centre of the vortex 

 is less dense on the whole than the air which is at the 

 .same temperature outside the vortex, then, since the 

 .nscending current must be compensated by a descend- 

 Insf current elsewhere, the nir will go through a 



NO. 2665, VOL. T06] 



thermodynamic cycle in which positive work will be 

 done at the expense of the heat communicated at the 

 earth's surface. 



The process may be compared rouchly to a Carnot's 

 cycle, in which the inflowing air at the'earth's surface 

 is isothermally raised in temperature, expands adia- 

 batically as it ascends, cools isothermally by radiation 

 at the higher levels as it flows outwards,' ani contracts 

 adiabatically in descending again. The work done 

 would appear as increaswl vorti:al motion if the 

 conditions were favourable, and the mechanical forces 

 causing the motion would be due to the differences of 

 hydrostatic pressure within and without the cyclone. 



J. R. Cotter. 



Trinity College, Dublin, Novemb'?r 13. 



Molecular and Cositiical Magnetism. 



Recent researches on magnetism tend to suggest 

 that the negative electron may be a magneton or 

 unitary electromagnet as well as a unitary electric 

 charge, consisting, that is, of an anchor-ring' of nega- 

 tive electricity in rotation about its axis of symmetry. 

 Such a magneton would behave mechanically like a 

 gyroscope; magneto-gyroscopic effects have been 

 previously considered and observed in relation to ferro- 

 magnetic bodies on the assumption that the ferro- 

 magnetism is due to electrons in orbital motion as a 

 whole. Wider conclusions can be drawn, how- 

 ever, if the magneton hypothesis is adopted, and the 

 deductions are of importance, not only in the theorv 

 of atomic and crystal structure, but also in relation 

 to cosmical magnetism. The following notes describe 

 a few of the more important consequences ; a detailed 

 account of the theory and of some experiments de- 

 signed to test its validity will be published shortly. 



k magneton rotating with anv kind of matter will 

 tend to align its axis parallel with the axis of rotation. 

 Since the electricity of the magneton is negative, the 

 direction of magnetisation will be related to that cf 

 rotation, as is the direction of translation to that of 

 rotation in a left-handed screw. This is the right 

 direction in order to account for the general magnetic 

 fields^ of the earth and sun as due simply to their 

 rotation. The explanation of the observed magnitudes 

 of these fields seems to present no difficulty ; nurrfrrical 

 details will be given in the forthcoming paper, where 

 also a theory will be suggested to account "for the rapid 

 radial diminution of intensitv in the ^imV <;. tirr.il 

 magnetic field. 



Mr. S. J. Barnett has shown by delicate experiments 

 on ferro-magnetic botlies that they become magnetised 

 slightly on being rapidly rotated, and has propounded 

 a theor}' acrordinij to which such bodies should 

 acquire magnetisation of amount proportional to tne 

 angular velocity, the factor being a universal constant 

 depending on the ratio (mass/charge) for an electron. 

 The fields observed and calculated (on this theory) 

 agree as to order of magnitude, but are quite in- 

 adequate to account for solar and terrestrial mag- 

 netism. The theon', however, apart from the fact 

 that it is based on the hypothesis of electrons in 

 orbital motion, seems to require serious modification. 



On the present theory, magnetisation by rotation 

 should be shown hv din- and para-magneflc bodies 

 as well as by ferro-mngnetic substances, and the inten- 

 sity should be proportional to the angular velocity only 

 when the substance is in such a state that the con- 

 straints exerted on the magneton by neighbouring 

 nuclei and electrons are strirtiv elastic. In this case, 

 moreover, the factor of proportlonalitv will not be h 

 universal constant, but will vary with the nature of 

 the constraints, and In particular with temperature: 

 In hot bodies the Intensity of magnetisation should 



