February 23, 1922^ 



NATURE 



257 



Debye's formula with a frequency v = 654i/\/T. 

 ie molecular heat of nitrogen is very approximately 

 Jven by a molecule based on Bohr's theory, with a 

 equency given by the gyrostatic formula. — U. R. 

 Passivity and overpotential. (i) Where a 

 stal is corroded by a liquid yie'ding an insoluble 

 rrosion product, the latter may either cling to the 

 »tal, forming a thin protective covering (often in- 

 iible), or it may become dispersed in the solution ; 

 the latter case it will not seriously interfere with 

 rrosion. It is probably the relative values of the 

 Iterfacial tension between the metal, corrosion- 

 luct, and solution that determines which will 

 cur. (2) The activation of passive metals by 

 jflorides is related to the known peptising action of 

 jtallic hydroxides by chlorides ; the passivication by 

 ;ans of chromates is connected in part with the 

 :culating action of chromates. (3) The fact that 

 ;tals with basic oxides are rendered active by acids 

 kd passive by alkalis, whilst those with acidic oxides 

 jd to become passive in acids and active in alkalis, 

 )ws that the invisible protective layer is " of the 

 Iture of an oxide-film." But it seems wrong to 

 entify it with any oxide known in the massive 

 ite ; probably we have to deal with a layer of oxygen 

 )ms connecting the metal on one side to the solu- 

 on the other. (4) Likewise, at a cathodically 

 slarised electrode we probably have to deal with a 

 layer of hydrogen nuclei connecting the metal to the 

 liquid. The hydrogen is probably in a state inter- 

 mediate between the elementary and the ionic, and 

 by forming a link between metal and solution it serves 

 to decrease the energy of the interface. Overpotential 

 may be due to the energy needed to desorb the 

 li\drogen. — A. W. Porter : Note on the vapour-pressure 

 >f ternary mixtures. The equation proposed for 

 rnary mixtures in a previous paper is here applied 

 n the case of mixtures of toluene, carbon tetra- 

 chloride, and ethylene bromide,' and is found to be 

 satisfactory. 



Royal Meteorological Society, February 15. — Dr. C. 

 Chree, president, in the chair. — C. E. P. Brooks and 

 J. Glasspoole : The drought of 192 1. The general rain- 

 fall in England and Wales was the least in 192 1, so 

 far as can be ascertained, since 1788. Individual 

 long records indicated that over a considerable part 

 of the south-east of England 192 1 was the driest year 

 for at least a century and a half. The months of 

 ■ i2i were not individually so remarkable as was 

 -hown by a comparison with the driest months known 

 > have occurred in the British Isles generally. As 

 hown by a map of standard deviation of annual rain- 

 lall, 1881-1915, for the British Isles, the least fluctua- 

 tions of annual rainfall occurred along the coast in 

 the north-west, increasing to a maximum in the 

 south-east and centre of the land masses. Construct- 

 ing charts showing the distribution of barometric 

 pressure over the globe during and preceding each of 

 the great droughts, beginning with 1864, it is found 

 that the conditions which commonly prevail during 

 dry spells are high pressure over the British Isles, the 

 greatest deviation from normal being usually over 

 -' )uth-east England ; low pressure over the Arctic 

 H'gions, especially near Spitsbergen ; and, generally, 

 low pressure over the tropics. The first factor is 

 related to the eleven-year sun-spot cycle, occurring 

 most frequently two years after sun-spot minimum 

 and three or four years after sun-spot maximum, so 

 that it tends to recur every five or six years. Low 

 pressure over the Arctic is related to ice conditions, 

 and tends to recur every four or five years. Great 

 droughts occur only when both these factors are 

 favourable. With pressure low over the Arctic, two or 

 NO. 2730, VOL. IO9I 



three months' warning of a drought would be given 

 by the development of high pressure over northern 

 Russia. — T. Kobayasi : A cyclone which crossed the 

 Korean Peninsula and the variations of its polar front 

 The cyclone passed over a mountain range in Korea 

 on March 25, 1918. It induced a secondary on the 

 farther side of the range along the steering line, which 

 extended upward until it joined with the primary ; the 

 secondary gradually grew stronger, and the original 

 centre disappeared. The existence of the polar front 

 was very distinctly marked, but complicated in 

 character. There were two or more squall lines 

 for one steering line. 



Edinburgh. 



Royal Society, February 6. — Prof. F. O. Bower, 

 president, in the chair. — J. M'Lean Thompson : The 

 floral structure of Napoleona imperialis, Beauv. The 

 flower structure of this curious African plant of the 

 Myrtle family has remained a puzzle since the dis- 

 covery of the plant in 1786. The flowers possess inside 

 the corolla a series of petal-like growths, which it 

 was held were produced during descent by trans- 

 formation and replacement of cycles of stamens of a 

 myrtle type of flower. These petaloid growths and 

 the persisting cycle of stamens, which themselves are 

 now partly sterile, are associated with a massy 

 fleshy disc which surrounds the base of the style. 

 The disc, stamens, and petaloid growths in question 

 have now been shown from developmental study to 

 have a common origin from a ring-like outgrowth 

 which normally in the Myrtle family bears numerous 

 groups of stamens, and were held to be the results 

 of replacement of stamens during descent. — G. W^ 

 Tyrrell: The pre-Devonian basement complex of 

 Central Spitsbergen. The rocks described constitute 

 the basement which underlies the flat-lying Devonian 

 and Carboniferous sediments in the region about the 

 head of Klaas Billen Bay. Thev form the southern 

 continuation of the extensive Wijde Bay region of 

 Urgebirge, where these ancient rocks begin to be 

 covered unconformablv by a sedimentary mantle. 

 Lithologically, they fall 'into a western zone of 

 "Archaean " facies,' consisting of quartz-schists, gar- 

 netiferous mica-schists, hornblende-schists and 

 gneisses, lit-par-lit-gneisses, and augen-gneisses, with 

 crvstalline limestones ; and an eastern zone consisting 

 of slates, quartzites, and limestones similar to those 

 of the Hecla Hoek svstem, the type-locality of which 

 (Hecla Hoek, in Treurenberg Bay) lies exactly on the 

 northern continuation of the line of strike of these 

 rocks. Hence this group is believed to be of Hecla 

 Hoek (probably Ordovician) age. The western schists 

 and gneisses, while showing great similarities to 

 rocks involved with the Hecla Hoek of the north- 

 western mountains, may be much older^ possibly even 

 Archaean. 



Manchester. 



Literary and Philosophical Society November i, 192 1. 

 —Mr. T. A. Coward, president, in the chair.— S. J. 

 Hickson : Some early autographs of John Dalton. 

 Variations in Dalton's handwriting were examined 

 and attention was directed to certain family records. 

 — R. W. James : The distribution of the electrons in 

 atoms. When X-ravs fall on an atom each electron 

 of the atom probably becomes a source of scattered 

 X-radiation ; the waves scattered by the electrons in 

 the direction of the incident light will be in phase, 

 and the total amplitude scattered in this direction will 

 be proportional to the number of scattering electrons. 

 If the electrons in the atom lie at distances from 

 the nucleus comparable with .the wave-length of the 

 X-ravs, the waves scattered from the different 



