422 HORARY VARIATIONS OP THE BAROMETER 



The amplitude of tbe diurnal oscillations of the barometer Avill diminish iu 

 m-oportion as we advance towards the higher latitudes; it will very nearly 

 vanish in countries where the sun no longer appears above the horizon, as in 

 those also on which it constantly shines during six consecutive months. 



The amplitude will diminisli when the sky is overcast or cloudy, but the oscil- 

 lations would not on that account cease altogether; for if fogs or clouds 

 intercept the view of the sun from the earth, they still receive the rays of that 

 luminary on their upper surface, Avhicli must consequently have its temperature 

 raised, as would be that also of the air which touches it, though in a less degree 

 than is the case when the rays of the sun strike the earth itself. 



Beyond the zone for which the length and heat of the days are, so to speak, 

 invariable, it should be in the warmest month of the year that the diurnal baro- 

 metric oscillation is most considerable : hence we infer that there is a maximum 

 for the year. But there is difficulty in admitt'ng a monthly maximum or mini- 

 mum. The existence of the latter, if observation should verify it in certain 

 localities, must depend on special circumstances, from which it will be easy, we 

 think, to disengage the phenomenon itself. As this phenomenon, however, ac- 

 cording to what we have so often said, is in an especial manner connected with 

 the temperatTire, there would be no positive ground for surprise if the months 

 which comprise the spring and sum.mer Avere found to have each a minimum at 

 the commencement and a maximum at the end ; while, for the months constitu- 

 ting the autumn and winter, there should be a maximum in the first days and 

 a minimum in the last days of each of those months. But, in our high latitudes, 

 and with all the causes which occur to derange the regularity of the course of 

 our seasons and their temperatures, it will, perhaps, be very difficult to establish 

 by direct observations what I am disposed to call the differentials of the phe- 

 nomenon of barometric diurnal oscillations. 



Permit me, in conclusion, to advert to a great variance which exists between 

 us. In several passages of your work 3'ou speak of variations in the weight of 

 the atmosphere, and you seem inclined to connect these variations of weight 

 with the barometric diurnal oscillations ; I might cite specially the last para- 

 graph of page 32, in which you say: " We have inquired how it is that, in the 

 same place, the weight of the atmosphere varies in the day according to the 

 hours, and in the year according to the months." For a long time I have been 

 opposed to this mode of interpreting the horary variations of the barometer; 

 these, in my opinion, are wholly independent of the proper weight of the atmo- 

 sphere, so that, on the supposition of the weight of the column of air remaining 

 absolutely the same for several months, the diurnal variations would not the 

 less be produced and within the same limits. They are the effect of the aug- 

 mentation or diminution of the volume and elasticity of the air, in consequence 

 of the heating or cooling of that air in contact with the surface of the earth ; but 

 this has nothing in common with the weight of the air simply as a gravitating 

 .body. I give but little credence to waves produced at the upper limit of our 

 atmosphere ; but had these Avaves several kilometres of altitude the phenomenon 

 of horary variations would not be materially affected ; they might slightly mask 

 it, perhaps, but would neither deprive it of its form or its importance. 



It is this unfortunate confusion of the weight and the action or elasticity of 

 the air, as also of the weight and elasticity of the watery vapor diffused in the 

 atmosphere, which renders the language of the baiometer so critical a matter to 

 interpret. This instrument, as regards the indications which it gives us and 

 the consequences which we draw from them, is much more a manometer than 

 a baroscope, as, at its origin, it was sometimes called. I hope to be able to 

 recur to this subject hereafter, and in the mean time conclude by saying that, in 

 the computation of heights by the barometer, there are circumstances in which 

 it will be necessary to take account of the horary variations, at the risk of 

 neglecting an eleiaeut of exactness. 



