July 25, 1878] 



NATURE 



129 



The storms generated in the Bay of Bengal afford unusual 

 facilities for studying the genesis of cyclones. We have obser- 

 vatories round three sides of the bay, and the sea is, at all times 

 of the year, traversed in all directions by numerous steamers and 

 saUing vessels, which have furnished abundant logs. Did 

 parallel currents play any important part in the production of 

 the vortices, they could not possibly escape our notice. But we 

 find that the antecedent conditions of a cyclone are light, 

 variable winds and calms, with a nearly uniform barometric 

 pressure all round the coasts ; and only to the south, in the 

 neighbourhood of the equator, is there any considerable move- 

 ment of the air, viz., from the west. Under these circumstances, 

 the pressure falls over some part of the bay ; most frequently 

 in the middle, and especially to the west of the Andamans. 

 This region of falling pressure is characterised by torrential 

 rains, with, at first, but little wind ; but after a day or two 

 (sometimes several days) of this weather, a cyclonic circulation is 

 set up, with a marked indraught in the neighbourhood of the 

 cyclone cradle, and thus the storm is generated. 



Having regard to these facts of observation, it appears to me 

 that it is in the condensation of the heavy rain (constantly noted 

 as "torrential" in ships' logs) over the cyclone cradle, that we 

 have the real source of the energy of the incipient storm. The 

 hypothesis of parallel currents fails to provide this energy ; for 

 it is obvious that the deviation of the winds under the influence 

 of the earth's rotation can furnish no energy, and can produce 

 only a moderate barometric depression, the amount of which 

 depends on the velocity of the original winds, and can be calcu- 

 lated by Ferrel's law. When this is reached, the system of 

 pressures and wind -movements will be in equilibrium. If (and 

 this I am not prepared to deny) a cyclone is sometimes gene- 

 rated between parallel currents, it must be that the energy of the 

 storm is supplied from some other source, and what this is, is, I 

 think, clearly indicated by the case of the Bay of Bengal 

 storms. 



It was first noticed by Mr. Eliot as a general fact, that, during 

 the formation of a cyclone in the Bay of Bengal, little or no rain 

 falls on the coasts ; while, as already remarked, it is exceedingly 

 heavy over the place of the storm's origin. The vapour gene- 

 rated over the bay, which, under other circumstances would be 

 carried away by the winds and condensed over the land, is then 

 condensed over the bay itself. The quantity of latent heat thus 

 set free is enormous ; and as Reye has shown, is ample for the 

 production of the most violent cyclone. It would be erroneous 

 to say that the air is thereby warmed and expanded, because, of 

 course, the very fact of its vapour being condensei proves that 

 it must be cooling ; but Welsh's and Glaisher's balloon observa- 

 tions show that in a cloud-laden atmosphere, the vertical decrement 

 of temperature is slow, as compared with that in a clear atmo- 

 sphere ; and the same fact is further illustrated by the temperature 

 of hill-stations in the wetter parts of the Himalaya as compared 

 with that of the plains at their foot. At Darjiling, for instance, 

 the temperature from June to August (the season of greatest 

 cloudiness and heaviest rainfall) is only 17° or 18° below that of 

 Goalpara. In February and March (the driest months) it is 

 between 23° and 24°. The explanation of these facts is afforded 

 by the different rates of cooling experienced by saturated and 

 unsaturated air, respectively, in an ascending mass of air which is 

 expanding under a constantly diminishing pressure. Saturated, 

 i.e., cloud-laden and rain-condensing air at 80° cools only 20° by 

 the work done during its ascent from the plane of 30 inches 

 pressure to that of 20 inches pressure, say through io,oco feet ; 

 whereas unsaturated air cools about 54° in the same ascent, the 

 exact amount varying slightly according to the quantity of vapour 

 it contains. The latent heat set free in the condensation of 

 cloud and rain is then entirely used up in the work of expanding 

 the cloud-laden air under a constantly diminishing pressure, and 

 economises more than half (indeed, in the case adduced, nearly 

 two- thirds) of the sensible heat which fiu-nishes the energy to 

 unsaturated air. Hence, an ascension'al current, however small, 

 once set in action in a nearly saturated atmosphere, such as exists 

 over the Bay of Bengal during the formation of a cyclone, 

 carries warm air to a greater height than in the clearer and drier 

 atmosphere around the coasts, relatively raising the mean tem- 

 perature of the former air-column, and of course reducing its 

 weight. This differential effect goes on increasing, and the 

 ascending current becomes more rapid, until the indraught 

 below attains the conditions of a cyclonic storm. 



Now, in the case of parallel currents, there must be between 

 them a region of calm ; and, if this is over a sea of high tem- 



perature, it is conceivable that, as in the Bay of Bengal, local 

 condensation may proceed for a sufficient time unchecked to 

 lead up to the formation of a cyclone ; but, in that case, the 

 cyclone will be generated, not immediately, as supposed by Mr. 

 Barham, by the energy of the pre-existing winds, but by their 

 affording conditions in which another and far more potent source 

 of energy comes into play. H. F. Blanford 



Dinard, France, July 10 



The Tasimeter and Magnetisation 



After perusing an accoimt, in a recent number of the 

 Scientific American, of Edison's Tasimeter, it occurred to one 

 of us to apply it to detect, and, if possible, to measure the elonga- 

 tion and shortening, which, as discovered by Joule, are pro- 

 duced in a bar of iron by magnetisation and demagnetisation. 

 Accordingly to test whether the effect could be observed in this 

 way, a rough specimen of the instrument was constructed, and 

 with it some preliminary experiments made, an account of which 

 may interest the readers of Nature. A small cylinder, about 

 half a centimetre in length and diameter, of the carbon used for 

 Bunsen's cells, rested with its ends which were slightly rounded, 

 in contact with two brass plates, one of which was fixed to a 

 rigid upright attached to one end of the base of the instrument, 

 Mhile the other, resting with one end on the base, formed a 

 spring, which in its normal position just touched the end of the 

 carbon. A coil containing four layers of insulated wire, six turns 

 to the layer, was wound round a tube ten centimetres long and 

 eight millimetres in diameter, and fixed with its axis in line with 

 that of the carbon cylinder, A piece of iron wire was then 

 placed in the axis of the tube with one end resting against the 

 spring, and the other in contact with the end of a screw working 

 in a nut fixed to a rigid upright at the end of the base remote 

 from the carbon. By means of this screw the pressure of the 

 iron bar on the spring, and consequently of the spring on the 

 carbon, could be varied at pleasure. 



A terminal of copper wire, was attached to each of the brass 

 plates bearing on the carbon, and joined up so that the carbon, 

 plates, and terminals formed one of the resistances of a Wheat- 

 stone's bridge, in connection with which a battery of one 

 Daniell's cell and a very delicate Thomson's reflecting galvano- 

 meter were used. When the iron wire forming the core of the 

 electro-magnet had been so adjusted that there was only a very 

 slight pressure on the carbon, the resistances of the bridge were 

 arranged to make the deflection of the galvanometer produced 

 by the ciurrent from the battery nearly zero. The galvanometer 

 and battery keys were then put down and the current allowed to 

 flow through the bridge during the remainder of the experiment. 

 The electro -magnet was then excited by the cxurent from three 

 of Thomson's Tray Daniells. This produced a deflection of the 

 image on the galvanometer scale of about fifty divisions in the 

 direction indicating a diminution of the carbon resistance, which 

 must have been caused by change of contact produced by 

 increased pressiu-e on the spring. The length of the iron core of 

 the electro -magnet had therefore been increased by magnetisa- 

 tion. When the magnetising force was removed the image im- 

 mediately returned to its former position. As a verification that 

 the diminution of resistance indicated by the bridge arrangement 

 was caused by elongation of the iron core, the adjusting 

 screw was turned forward through a very small distance, when 

 the deflection was found to be in the same direction as before. 

 When the screw was brought back the image on the scale returned 

 towards its zero. Experiments with various strengths of current 

 gave perfectly accordant results. 



We hope by replacing the comparatively rough adjusting screw 

 by a micrometer screw to be able to make some measurements of 

 the exact amounts of elongation or shortening produced in a 

 piece of soft iron or steel by given changes of magnetic intensity. 

 It may be remarked that this method of measurement could be 

 advantageously applied in cases where the amount of change of 

 dimensions to be discovered or measured is very small, but the 

 force whic"h it could be arranged to produce abundant. 



University of Glasgow, July 12 Andrew Gray 



Thomas Gray 



Physical Science for Artists 

 The curious phenomenon described by Prof. Briicke and Mr. 

 Norman Lockyer, under the name of "les rayons de crepus- 

 cule," though rare and uncommon in the island of Ceylon, is 



