650 



NA JURE 



[October 31, 1901 



water covers itself with drifting clouds of so-called steam. 

 There are doubtless surface fogs which correspond to 

 this condition, though when the evaporation is very rapid 

 there may be, as Mr. C. T. K. Wilson has shown, a clear 

 layer immediately in contact with the water surface. 

 Fogs which have their origin in this mixture of the rising 

 vapour with cooler air may be called " steaming water 

 fogs." On the other hand, if any surface is sufficiently 

 cold the absorption of heat from the air in contact with 

 it may cause condensation in the air close to the surface, 

 and fogs arising in this way may be called, for the sake 

 of brevity, " cold-surface fogs." Vigorous radiation such 

 as takes place on a clear night from grass may cause a 

 fog of this character, and in regard to the persistence 

 of a fog under these circumstances the transparency, or 

 rather translucence, of fog for radiation is an important 

 but not well-known factor. In these cases the source of 

 the water supply is easily identified as being the water 

 surface in the one case and the air in contact with the 

 surface in the other. 



Clouds, consistmg of more or less detached masses 

 of fog formed in either of these two ways, may drift 

 like the steam from a locomotive or a sea fog over land, 

 and a fog may thus visit a locality which has had no share 

 in supplying the water. There is, moreover, another pos- 

 sibility which may be connected with the question as to 

 why fogs are more prevalent in winter than in summer, 

 in spite of the fact that the store of moisture in the air is 

 larger and changes of temperature are more pronounced 

 in the warmer months. Rain has been defined as a 

 falling cloud which reaches the surface before the 

 evaporation of the globules is complete. The rate of 

 fall depends on the size of the particles, but in still atmo- 

 sphere even the smallest particles make their way down- 

 ward. In summer the falling cloud may consist of any- 

 thing Ijetween a thunder shower and drizzling rain. In 

 winter, when the supply of moisture is less and, over 

 towns, the supply of nuclei for condensation is greater, 

 the counterpart of the summer drizzle may be so light as to 

 be classed as fog or mist, and fall with extreme slowness. 

 In this case the water supply comes from strata above the 

 surface. There are certainly some fogs in which there 

 seems to be a gradual deposit of moisture on horizontal 

 surfaces, and not merely on specially cold surfaces. It 

 is true that winter fogs are often associated with high 

 barometric pressure, generally a fine weather associa- 

 tion, but under similar conditions of pressure very light 

 rainfalls on our eastern and northern coasts are some- 

 times experienced. Whether electrical conditions, which 

 are exceptional in foggy weather, may account for the 

 formation or accelerate the falling of the cloud in such 

 circumstances I cannot say. 



If we call this third form of fog, due either to the 

 surface drifting or the downward descent of a cloud formed 

 above the surface, a " cloud fog," we have altogether 

 three forms — "steaming water fogs," " cold-surface fogs" 

 and " cloud fogs." It is evident that, of these three, two 

 depend upon local conditions which may possibly be 

 identified, while the third is at least much more inde- 

 pendent of local conditions and its incidence may be as 

 capricious as the summer cloud. 



The consideration of the observations from this point 

 of view requires more than mere organisation. It in- 

 volves a special knowledge of the physics of the atmo- 

 sphere applied to observations of a somewhat special 

 kind, and may need some appropriate apparatus. It is 

 hoped that circumstances will allow the statistical in- 

 vestigation to be combined with the consideration of such 

 physical questions as those which I have indicated ; but 

 the time for arrangement is short, and it is possible that 

 the physical side of the investigation may have to wait 

 for a more faveurable opportunity. The primary con- 

 sideration at present is the suitable organisation of trust- 

 worthy observations. W. N. Sh.wv. 



NO. 1670, VOL. 64] 



ALUMINIUM AND ITS USES. 



T^HE number of metals available in large quantities for 

 -*■ industrial purposes is so very small, that the success- 

 ful introduction of a new one must be of the greatest 

 interest. The affinity of aluminium for oxygen much 

 exceeds that of iron at a red heat. Iron oxide is reduced 

 by carbon at that temperature, while alumina cannot be 

 reduced in this way except in the electric arc. And thus, 

 though the ores of aluminium are more widely distributed 

 even than those of iron, yet the former metal remained 

 for long unknown, and until lately was comparatively rare. 



About fifty years ago, the researches of Wohler and 

 Deville led to the latter's process for the production of 

 aluminium on a commercial scale, in which the vapour of 

 aluminium chloride was led over heated sodium. The 

 price of the new metal fell rapidly, but was ahvays high 

 and dependent upon that of sodium. 



In 1854, Bunsen and Deville showed independently 

 that aluminium could be obtained by electrolysis from a 

 bath containing the chlorides of both aluminium and 

 sodium in a state of fusion, the latter chloride acting 

 merely as a flux. But at that time the cost of electrical 

 energy was prohibitive. 



The first successful electrical process was that of 

 Cowles, in which alumina is reduced by carbon in the 

 electric furnace. It rapidly superseded the old chemical 

 method, in spite of the reduction in the price of sodium 

 by the Castner process, but had soon in its turn to give 

 place to the processes of Hall and Heroult discovered in 

 1886 (though not successfully worked until some years 

 later). In these, a bath of the fused fluorides of aluminium 

 and sodium is employed. They occur naturally combined 

 as cryolite ; and serve, when melted, as a solvent for 

 alumina, which by itself would, of course, be almost in- 

 fusible. The alumina is electrolysed by a current intro- 

 duced at a carbon anode, and further alumina is added 

 as the metal collects at the other pole. The bath must 

 be maintained at a red heat, and an electromotive force 

 of somewhere about five volts is needed. 



It is this process which has brought down the cost of 

 aluminium so much of late. Other methods, as those of 

 Blackmore and tiooch, depending upon the preparation 

 and subsequent electrolysis of fused aluminium sulphide, 

 are said to be yet more economical, the sulphide being 

 much more readily decomposed than the oxide. But the 

 saving in electrical energy does not yet seem to make up 

 for the greater expense of working materials. 



The cost of water-power, even in situations offering 

 great natural advantages, cannot be indefinitely reduced, 

 owing to the great capital outlay needed for hydraulic 

 works. There is, then, little doubt that the present cost 

 of aluminium, about eighteenpence a pound, represents 

 roughly the lowest figure at which the Hall and Heroult 

 processes can profitably be worked. With a density 30 

 per cent., and a conductivity 60 per cent, that of copper, 

 pure aluminium conductors can transmit the same elec- 

 trical energy over a given distance with only half the 

 weight of metal. As an electrical conductor, therefore,, 

 aluminium at eighteenpence is equivalent to copper at 

 ninepence a pound, or 84/. a ton, a figure considerably 

 below what it lately reached. 



It is curious to observe how entirely dependent the 

 electrical engineering industry is upon the price and the 

 conductivity of copper. The former largely determines the 

 degree of success, or at all events the method of carry- 

 ing out, of electrical power transmission schemes ; while 

 the latter, in conjunction with the permeability of iron, 

 actually decides the scale upon which our electrical 

 machinery must be built, since the output or eft'ort for a 

 given speed of running is always limited by the heating 

 which occurs ; and this, at full load, arises mainly from 

 the imperfect electrical conductivity of copper. 



Aluminium offers, it is true, no prospect of reduced 



