244 On the Normal Rainfall of Bengal. [No. 3, 



It must be promised that very few of the registers can pretend 

 to accuracy, and as will be seen from the figures indicating the 

 number of years from which each monthly average has been com- 

 puted, very few are complete for the entire series of years. It is clear 

 from the character of the original records, that the value of the re- 

 gister in each case has been determined very much by the amount of 

 interest taken in it, or the supervision that could be exercised over 

 it by the local officer, and in some cases it would appear to have 

 been treated in a very perfunctory manner. In some cases, the 

 register has been discontinued for several years consecutively, in 

 others for three or four months only, e. (/., while the rain-gauge was 

 sent to Calcutta for repair, and some sudder stations appear never to 

 have been furnished with rain-gauges. I have omitted many stations, 

 the data of which are generally doubtful, or insufficient to furnish 

 a fair average result, especially those in which the earlier series 

 shows a marked discrepancy with the later. On the other hand, 



1 have admitted one or two registers presenting points of special 

 interest, and which I have reason to believe trustworthy, although 

 extending over but a short period. "What kind of gauge may 

 have been used in the earlier years I am unable to say ; of late years, 

 the form commonly in use is that which consists of a deep narrow 

 receiver, in which moves a float cariying a graduated brass rod. 

 The rise of the float is read off on the rod at its intersection with 

 a bar which crosses the mouth and the funnel and through a hole 

 in which the rod slides.* Gauges of this kind in unpractised or 

 careless hands are subject to error in many ways ; the general 

 tendency of which is, that the quantities indicated are less than the 

 actual rainfall. This I am disposed to believe is very generally the 

 case with the registers here summarized, to the extent perhaps of 



2 or 3 per cent, of the total rainfall, but any such estimate must 

 necessarily be very vague. 



I have classified the stations in groups according to the chief 

 physical divisions of the country, and their exposure to the vapour- 

 bearing winds. Mr. D o v e in his well known treatise on the Eain- 

 fall of the torrid zonef has classified the Bengal stations in two 



* This form is figured as No. II in the Second Report of the Rainfall Com- 

 mittee of the British Association. Brit. Assoc. Rep., 1867, Plate IV. 



f Klimatologische Beitrage, Vol. I. Ueber die Vertheilung des Regens auf 

 der Oberflache der Erde. Erster, Theil. Die Regen der Heissen Zone. 





