124 Irregularities of Atmospheric pressure in Bengal 8fc, [No. 2, 



by errors arising from faulty position or manipulation of the instru- 

 ments. Accordingly I took every means in my power to verify the 

 elevations, and either by personal inspection or otherwise, to satisfy 

 myself that the instruments were properly placed and observed. 

 The barometers had all been compared with the Calcutta standard 

 before they were issued, and the errors thus ascertained had been 

 applied as corrections to their readings. At stations that I visited,* I 

 repeated the comparison with one or two mountain barometers which 

 I carried with me, and which had been compared with the standard 

 before my departure and were again compared on my return. In no 

 case did the result of the second comparison differ from that of the 

 first by more than a quite trivial amount. Some of the stations, f 

 moreover, have been supplied with duplicate barometers since the 

 peculiarities above noticed first attracted my attention, and in these 

 cases, a comparison was made between the two instruments as soon 

 as possible at the station, and their recorded errors thus made to 

 furnish a check on each other. I mention these details because in 

 this country the barometric variations are so small in comparison 

 with those in Western Europe, that it is of the utmost importance 

 in order that the conclusions based on their readings may be trust- 

 worthy, that all merely instrumental errors be most carefully 

 eliminated. Any confidence that my facts may lay claim to, will 

 depend on the assurance that all ascertainable causes of error have 

 been carefully ascertained and allowed for. 



These precautions then having been duly observed, and not 

 having afforded any explanation of the observed anomalies, J the 

 conclusion became legitimate, that they were real atmospheric phe- 

 nomena and not apparent and instrumental only : and this conclu- 

 sion was confirmed by the fact, that in some cases the same pecu- 

 liarity was shewn by two or more neighbouring stations. Finally 

 during the last cold weather (1868-69) I observed that certain 

 stations which, during the S. W. monsoon, had shewn an excessive 

 barometric depression, now exhibited an opposite tendency, an 

 excess of atmospheric pressure ; and that this like the former pecu- 



* Dacca, Chittagong, Shillong and Monghyr. 

 + Saugor Island, Cuttack and Akyab. 



X With one exception. The elevation of Chittagong had been erroneously 

 reported, as shewn by my verification. 



