2G0 REPORT — 1873. 



The experimental gauges erected some years since at Calne, at the ex- 

 pense of Col. Ward, and subsequently removed to Strathfield Turgiss and 

 Hawsker (and of which the results were reduced, presented to this Com- 

 mittee, and by them inserted in their 1869 and 1870 Reports), have been 

 finally dismounted and preserved for future use if required, it being consi- 

 dered that the doubtful points which they were constructed to test have 

 been thoroughly settled. 



During the decennial period, extending from 9 a.m. January 1st, 1860, to 

 the same hour on January 1st, 1870, there were 317 records of rainfall kept 

 in the British Isles, without the omission of a single shower. These records 

 therefore give 38,040 monthly values, or 3170 values for each month of the 

 year, and afford by far the most reliable basis for investigation into the 

 seasonal distribution of rainfall ever yet available. Accordingly your Com- 

 mittee have had them all converted into percentages of the yearly totals at 

 the several stations, and tabulated in the same manner as those for previous 

 decades given in our Report for 1868. We give on the present occasion in 

 Table I. the percentages for each individual station, because it has been 

 remarked that we have not given monthly averages, and these percentages 

 afford the means of readily obtaining such averages. It is merely necessary to 

 shift the decimal point two places to the left to convert the percentage into 

 a factor for deducing the monthly amount from the mean annual amount 

 given in the column preceding the monthly percentages. Por example, the first 

 station is Shrewsbury, of which the mean annual amount was 19-499, and 

 the January percentage 8-6, which by shifting the decimal point is converted 

 into the factor -086, and 19-499 X •086=1-677 in., the computed January 

 fall. The true January mean at Shrewsbury is 1-675 in. ; and although 

 the mean, computed by the above method, would not in all cases be in 

 such remarkably close agreement with the true mean, the difference would 

 never be of any consequence. 



In Table II. we give the means for each group, and, for comparison, the 

 corresponding values for the previous decade 1850-59, and also the depar- 

 tures of each group from the mean of each district. These values strengthen 

 the evidence which we adduced in our 1868 Report of the greater relative 

 wetness of winter months at western stations, and especially at those of 

 large rainfall. But though they corroborate the fact of the oscillation, they 

 rather reduce its amount. For instance, at western stations in England we 

 have the following monthly percentages for stations at which the average 

 is 20 to 25 in. :— 



1850-59. 1860-69. 



January 7*9 January 7-8 



July 10-6 July 83 



Difference . . 2-7 Difference . . 0-5 

 GO to 65 in. :— 



January 13-9 January 11-2 



July 7-4 July 5-4 



Difference . . 6-5 Difference . , 5-8 



It is satisfactory to find that the general inferences drawn by Mr. Gaster, 

 and quoted in our 1868 Report, are so far corroborated by the fuller in- 

 formation now obtained — that, except as hereinafter noted, we may refer to 

 that Report as giving a fair resumi of the facts in the present, always re- 



