ON THE RAINFALL OF THE BRITISH ISLES. 199 



2. What has been done since 1860. — Having thus described what had been 

 done previously to 1860, I proceed to mention briefly what has been done 

 since then ; and if it seems that this portion is egotistical, I reply I cannot 

 help it : it would be strange indeed if the one person who has devoted him- 

 self solely to the subject were not mixed up with most of its ramifications. 



In the first place, then, it has already been stated that prior to 1860 

 there had been no general collection of all reliable rainfall-records ; this 

 was the primary object I had in view, knowing that without such a col- 

 lection it was impossible to determine, with any approach to accuracy, 

 either the geographical distribution or secular variation of British rainfall. 

 To what an enormous magnitude this collection has grown may be inferred 

 from the Tables attached to this Report ; but as a ready means of enabling 

 engineers and others to estimate the stores awaiting them, I may say that, 

 supposing the lists closed at once, I believe the publication of the monthly 

 values and discussion of the observations would fill two of the bulky annual 

 volumes of this Association. Of course bulk is a poor test of value ; but I 

 wished to give some better description than that of " twenty-three folio 

 volumes of MS.," which are already full. I should not omit to state whence 

 these observations have been copied. Most of them have been obtained from 

 the observers, many of whom have sent me registers of ten, twenty, thirty, 

 and even more years in length ; some are from privately printed papers, 

 others from magazines, scientific journals, and Transactions of the Royal and 

 other Societies. Particulars as to the position of the gauge are always care- 

 fully noted along with the observations, when they can be obtained. I had 

 not been at this work very long before I felt the desirability of publishing 

 some information of a reliable character, and at the close of 1860 I published 

 a Table giving the total depth in that year at 168 stations, being more than 

 double the number ever previously collected ; this Table has grown into an 

 annual volume, both by giving additional information on the undermentioned 

 and other subsidiary investigations, and by its increased body of contributors, 

 now numbering over 1000. 



Next to compiling the general Tables and annual reports, the most trouble- 

 some, but probably also the most important, work is the examination of the 

 gauges actually in use, both as to their own accuracy and also as to the suita- 

 bility of their position. I need not point out how necessary this is ; it must 

 be evident to everybody that until it is done there will be liability to all sorts 

 of errors in deducing results from observations with gauges either themselves 

 incorrect, or badly placed. I do not wish to be discursive ; or I coidd quote 

 man} - positions in which I have found gauges, nothing less than absurd. (Very 

 gross errors are now impossible, owing to the number of new tested gauges 

 interspersed among the old ones and available as checks upon them.) This 

 examination of the gauges in situ involves an amount of travelling which takes 

 far too much time and too much money for me to make any very great progress 

 with it. However, I have visited 113 ; and if this Association will find the 

 needful funds, I will endeavour to double the number before I draw up the 

 next annual Report. 



Next to personally visiting each gauge, the best plan is to test the gauges 

 before the opticians send them off to their destinations. Of the exact number 

 thus tested I have no record, but believe it to be about 300, a number easily 

 mentioned, but by no means so easily examined. 



At the outset of the investigation an attempt was made to collect returns 

 of the number of rainy days or days of rain ; but it was speedily found that 

 utter want of uniformity prevailed, and that it was useless to attempt to do 



