1843.] in Barometric Observations. 297 



ward, which is specifically lighter, until the air here being similarly 

 cooled in like manner, proceeds to press on and displace the warmer 

 air to the westward. When the influx balances the efflux, there is the 

 evening maximum : after which the accumulated air gradually dis- 

 perses itself till towards morning, when it is again pressed on before 

 sunrise by the air heated in the east. 



If this be a true account of the atmospheric tides, it is plain, that 

 supposing the air to be calm, we may expect the disturbing causes 

 will be least about the time of the morning minimum; and that at 

 any time during the night if the air be calm, these are likely to be 

 much less influential than during the day, for then they are entirely 

 free from the direct action of the sun, which evidently has a great 

 effect on barometric heights by the inequalities of temperatures which 

 it occasions in different places. We know that even in the hottest 

 weather, the temperature on the Sea Coast is tolerably uniform when 

 compared with that of places inland, particularly when these inland 

 places are on elevated plains. In such cases we cannot safely assume 

 that the mean of the temperatures at the two stations will truly re- 

 present that of the intermediate column of air: or perhaps, to speak 

 more properly, it cannot be assumed that the temperature and mois- 

 ture at the upper station will approximate to those of a place on the 

 same level immediately over the lower station ; besides which, the 

 barometric pressure at the inland station may be very different from 

 that at the supposed station on the same level if, as is most likely, the 

 state of the wind should be different in the two places. In short, any 

 thing analogous to wind or current which would affect the correctness 

 of water-levelling may be expected in a still greater degree to affect 

 the correctness of barometric-levelling: and we may infer as a gene- 

 ral conclusion that, besides the goodness of the instruments employed, 

 the trustworthiness of barometric measurements will greatly depend 

 on the ca«re and skill with which the observers avoid the influence of 

 disturbing causes. 



lUh April, 1842. 



2 R 



