April 14, 1 881] 



NATURE 



559 



studies of his University: "To explain difficulties in 

 these questions, the atmospheric strata have been shuffled 

 in accordance with laboratory experiment-;." Thus, for 

 example, the mean pressure of the atmosphere remains, 

 on the average of the whole year, o'ojS inch lower for 

 every 100 miles we proceed north in this country, a differ- 

 ence which is called ^gradient, as if it were a fall on a 

 railway line, though it is really the position of equilibrium 

 like that of the watery ocean, which has also a gradient 

 of nearly thirty miles from the equator to the poles. 



I have previously remarked that at Bombay the maxi- 

 mum pressure precedes nearly by a month the minimum 

 temperature, while the minimum pressure is a month 

 later than the maximum temperature. This is also true 

 for all the stations in North India. At Madras, however, 

 and Trevandrum, January becomes the month of maxi- 

 mum pressure. I do not, therefore, place much weight 

 on this fact as showing that the two oscillations are not 

 cause and effect. The month of maximum pressure at 

 Pekin agrees most nearly with that of minimum tem- 

 perature. 



I have stated in the first article on this subject that I 

 did not admit that the oscillation of pressure was due to 

 that of temperature, and therefore could not allow that a 

 higher annual mean temperature [would in any case cause 

 a lower annual mean pressure] From the fact that the 

 annual variation of pressure and temperature in Central 

 Asia is greater than in any other portion of the globe, the 

 greatest pressure coinciding nearly with the lowest tem- 

 perature, and the least pressure with the highest tempera- 

 ture, it was concluded by Mr. Chambers that years of 

 gi'eatest mean pressure should also be years of least mean 

 temperature. Now if we assume that the pressure depends 

 only on the mass of the air and watery vapour in it, as the 

 former is constant, and the latter, the only variable part, is 

 greatest when the temperature is highest, it would follow 

 that years of greatest heat should be years of greatest 

 pressure, which is just the reverse of the conclusion 

 deduced by the analogy from the annual variations. 



Indeed, it is one of the great difficulties in the hypo- 

 theses which have been proposed, to explain the annual 

 variation of pressure of the mixed atmosphere, that when 

 we subduct the vapour pressure, as far as our means of 

 calculating this exist, we have a much larger dry air 

 oscillation than before. 



I gave, however, different reasons for concluding that 

 the range of temperature w-as not itself the cause of the 

 diminished pressure, although the two go nearly together. 

 One was that the observations of Bombay showed the 

 greatest pressure to precede the lowest temperature by a 

 month, and this is true for all the stations in the groups 

 of North India already given. I also pointed out that 

 were the two directly related, the mean pressure at Tre- 

 vandrum should be greater than at Pallamcottah by nearly 

 one-tenth of an inch, which is not the case, the isobars 

 and isotherms having no relation to each other. 



If we suppose that we have the same atmosphere over 

 each station as over the whole earth, there is no possi- 

 bility of explaining the variation of pressure by that of 

 temperature. The only known property of heat which 

 affects the mass has no doubt been employed to cause 

 the hotter air to How away somewhere, and surely in that 

 case it should flow to the nearest colder station, where 

 the pressure is less ; but we have seen that this is not so 

 in the case of Trevandrum and Pallamcottah, nor is it so 

 in the valleys of the Ganges and Indus, where the oscil- 

 lation increases as we ascend from the sea. These oscil- 

 lations proceed with the greatest regularitj', approximately 

 in proportion to the temperature variation from month 

 to month, and without the slightest regard to the hypo- 

 thesis which should cause equilibrium in twenty-four 

 hours, by the sliding of the most expanded masses over 

 those least so. In what way, then, can we associate 

 the two oscillations if one is not the cause of the other ? 



I have long ago suggested that the varying humility of 

 the air may be in question ; this is only a suggestion. 

 I do not mean the mere tension of vapour — as already 

 stated when we try to get rid of that from the total 

 atmospheric pressure, the subject becomes more difficult, 

 the dry air oscillation being greater not less than that of 

 the whole — but if we suppose that the attraction of gravity 

 is not the only attraction which affects the pressure of the 

 atmosphere, but that this pres-ure varies through some 

 other attracting force such as an electric attraction of the 

 sun depending upon the varying humidity of the air, and 

 this a,i,'ain depending on its temperature ; we should find a 

 method of relating the two variations which does -not exist 

 if gravitation alone is employed. 



It is quite certain that many physicists will not admit 

 the idea of an electric attraction on our atmosphere in the 

 present state of our knowledge, hence the efforts to make 

 expansion, and a shuffling of the atmospheric strata 

 suffice. We must not, however, in our ignorance, 

 attempt to force conclusions in opposition to facts, and if 

 these can be satisfied more easily and with greater 

 probabilities in its favour by the aid of the hypothesis of 

 an electric attraction of the sun, that hypothesis will have 

 a better claim to acceptation than the other. 



I shall here note a few facts which cannot be explained 

 by thermic actions. 



1. I have shown that on the average of many years' 

 observation in our latitudes the mean pressure diminishes 

 at the rate of o'03S inch of mercury for every one hundred 

 geographical miles we proceed towards the north. This 

 has been called a gra-lient fro n the similar term used in 

 railway slopes ; but it is no slope, it is a level of a surface 

 of equilibrium like that of the sea. It is the mean heights 

 of the b.arometer at the sea-level which indicate the form, 

 if we may say so, of the equilibrating atmosphere. 



2. In India we have seen that the atmospheric pressure 

 oscillates at each station even when these are quite near 

 to each other, independently of the known laws of equi- 

 libriun of pressure of gases. 



When we turn to the semi-diurnal oscillation of the 

 barometer we are only amused at the attempts made to 

 explain it by shuffling the atmospheric strata. Nothing 

 can be more certain than that the theories of expansion, or 

 resistance to expansion and overflow, are the vain efforts 

 to make the laws of nature agree with a theory. Over the 

 great ocean within the tropics, where the diurnal varia- 

 tions of temperature are small and the air is absolutely 

 without perceptible currents for days together, the baro- 

 meter rises and falls a tenth of an inch twice in twenty- 

 four hours with the regularity of the solar clock. The 

 action of the sun on the whole atmosphere which pro- 

 duces this movement varies chiefly during the day hours 

 at inland stations with the temperature oscillation, so 

 that, as in the case of the annual variation, the fall of the 

 barometer at 4 P.M. is greater in the same latitude as the 

 temperature is higher. This variation occurs during the 

 most complete calms ; the smoke rises vertically from the 

 plains of Tinnevelly ; no current is visible in the motion 

 of the clouds ; yet the barometer falls at four in the 

 afternoon as it did at four in the morning, only it falls 

 farther. 



THE ETNA OBSERVATORY 



THE accompanying illustration of the Observatory on 

 Mount Etna is reproduced from the Memoirs of the 

 Italian Spectroscopic Societ,v. It shows that the building 

 is so far complete, and surmounted by its revolving dome 

 for the protection of the large Merz equatorial of thirty- 

 five centimetres aperture. In the engraving the volcanic 

 cone appears much nearer the Observatory than it really 

 is. The work of building was suspended during the 

 stormy weather of 1879, but was completed in the 

 summer of last year. But it cannot be said that the 



