CORRESPONDING SOOIETIES. 443 



then, it is apt to be mistaken for a geographical variation. The only method of 

 which I am aware, which has given completely satisfactory results in detecting 

 faulty records, is the cartographical method elaborated by Dr. Mill and used 

 largely by the British Rainfall Organisation. In the course of ordinary routine, 

 we have constructed maps, mostly based upon several thousand records, for 

 individual days, months, years, and for the average of a long period of years. 

 As I have shown in my paper on ' The Relation of Rainfall to Configuration,' ■* 

 an individual day's rainfall may, or may not, according to its origin, show any 

 lonformation to the land configuration; a month's rainfall is nearly always 

 more or less dependent on the configuration (a winter month's rainfall always) ; 

 whilst the total for a year, or the average for a period of years, is always 

 very intimately related. In drawing rainfall maps it is common to find certain 

 records exhibiting a want of harmony with others in this respect. Sometimes 

 one record will appear out of harmony month after month. The range of 

 variation which can be safely overlooked as fortuitous diminishes with increase 

 of period. Thus in a single month a variation of, say, 25 per cent, might be 

 due to some local thunderstorm, and in a month when thunderstorms are known 

 to have occurred it would not be safe to assume that an even larger variation 

 was due to error. In a map of the rainfall of a year a variation of 10 per cent., 

 not explained by the configuration, should give rise to suspicion, and in the 

 case of a map of average rainfall for thirty or forty years a 5 per cent, variation 

 would almost certainly indicate error. In a very large number of cases records 

 showing want of harmony in this way have been made the subject of special 

 investigation. In practically every case the gauge has been found to be faulty, 

 either in construction or exposure in one of the ways described. It is true 

 that a very small number of cases of persistent variation remain unexplained, 

 and that the subject is far from fully investigated, but from a body of experi- 

 ence gained during about twenty years devoted to the subject, in co-operation 

 with Dr. Mill, I have confidence in putting forward the opinions expressed in 

 this paper. 



The President pointed out that in Hampshire there was a marked difference 

 between the rainfall on the sea coast and that a few miles inland, the cool winds 

 blowing from the sea keeping away thunderstorme coming from the north and 

 during the summer months north-west. 



The Rev J. 0. Bevan (Woolhope Naturalists' Field Club) said that local 

 Societies should be encouraged to increase the number of rain gauges. 



Dr. J. S. Owens (Honorary Secretary to the Advisory Committee on Air Pol- 

 lution) said that Mr. Salter referred to the need for increasing the number of 

 observing st.ations for measuring rainfall. The Advisory Committee on Atmo- 

 spheric Pollution experienced the same need for a multiplicity of stations for 

 measuring deposit of impurities. It appeared therefore that now would be a 

 suitable moment for considering! whether the rain gauges and deposit gauges 

 could be combined in one instrument. This suggestion wae made by Professor 

 Gee and is well worth careful examination, especially as now both the rainfall 

 and deposit of impurity measurements came under the control of the Meteor- 

 ological Office. The deposit gauge now used was really of the nature of a large 

 rain gauge with four square feet catchment area, the large size being necessary 

 if any estimate of the proportion of tar and ammonia present is to be made; 

 useful information conld be obtained from a smaller and cheaper gauge which 

 would receive the rainfall and impurities ; the latter might then be separated 

 into soluble and insoluble matter instead of making: a more elaborate analysis. 

 His object in speaking was to suggest that the possibility of combining the' two 

 measurements of rain and impurity should be "carefully gwne into before new 

 stations were set up. 



The President mentioned that in India at a height of seven or eight thousand 

 feet the first rain often broucfht down a muddv deposit. 



Mr. Wilson L. Fox (Royal Polytechnic Societv of Cornwall) asked how the 

 'measurement of the soluble and insoluble matter could be carried out in prac- 

 tice by amateur rninfall recorders. 



fiofessor W. W. 'Haldane Gee .=aid that a serious demand came from hydro- 



* Proc. Inst. Water Engineers, vol. xxiii., 1918. pp. 45-91. 



