20 



niscoviinv 



Rainfall is not only influenced by geographical 

 position, hut also by the height and proximity of 

 mountains and of the sea, in some cases even of hills 

 and forests. If, therefore, we have a neighbour living 



FIG. I. — SNOWDCIN RAIN-GAUGE 

 AND MEASURIv GLASS. 



on a hill near us who keeps a record of the rainfall, 

 that is not a sufficient reason for our omitting measuring 

 the rainfall by a rain-gauge in our own back garden. 

 In partial storms it is quite possible that we record 

 more than an inch of rain, and our neighbour's gauge 

 remains perfectly dry. The opposite may also be 

 the case. 



The instruments described in these articles will be 

 those used at — 



(i) Stations of the First Order, or Self-recording 

 Observatories. 



(2) Stations of the Second Order, or Normal 



Climatological Stations. 



(3) Stations of the Third Order, or Au.xiliary 



Climatological Stations. 



(4) Telegraphic Reporting Stations. 



(5) Anemograph_Stations. 



The descriptions, however, will be given not in the 

 order in which they are used at these stations, but 

 in the order of simplicity in use, commencing with 

 the simplest. Each instrument will be allocated 

 its proper position at each station in a summary in 

 the concluding article. 



The best and most usual pattern of rain-gauge in 

 use to-day is the " Snowdon," a gauge having a funnel 

 5 inches (127 centimetres) in diameter. 



The funnel has a brass rim at the top, the inner 

 edge of which is cut sharply to the diameter mentioned. 

 The funnel itself is 4 inches in depth.' 



The reason for adopting this gauge with a deep 

 funnel is that snow cannot blow or hail jump out of 

 it after it has been caught. In old-fashioned rain- 



' Expcriment.s made recently bv Mr. W. H. Dines at Benson. 

 O.xford. seem to suggest that the ideal rim should have a slope 

 inside as well as outside, the cross-section bcmg nke the 

 gable-end of a house. — Meteorological Maga:i)ie, June 1920 



gauges with shallow funnels, relatively a large amoui. 

 of snow was lost during high winds. In consequen< 

 the amount of precipitation appeared too little whci 

 measured in these gauges. 



A second reliable gauge is the Meteorological Offi. 

 pattern rain-gauge. The diameter of the funnel <■: 

 this one is 8 inches {203 centimetres), and except 

 in size is similar in every way to the Snowdon. 

 Observations made with either of these instrument 

 are accepted as equally trustworthy by the Meteon. 

 logical Office and the British Rainfall Organisation. 

 In purchasing rain-gauges, a prospecti%'e bu3-er will 

 find it best to get a copper apparatus. A coppt! 

 rain-gauge costs more initially, but it lasts practicaih 

 a life-time, and it does not require periodical repaintin:. 

 as the cheaper ones do. Secondly, he should decid' 

 whether he intends to measure the amount of precipi 

 tation in inches or in millimetres. It makes n< 

 difference in which units the measurements are mad- 

 provided one unit is kept to throughout. An inch i- 

 equal to 254 millimetres, and a millimetre is 003937 

 inch. For convenience, the millimetre may be con- 

 sidered equal to 0*04 inch, but it is clear that if this 

 value be used for converting from one unit into the 

 other, small errors arise, while if the correct value be 

 used, it involves quite a lot of arithmetic. It is best, 

 therefore, to keep to one unit throughout. 



If the rainfall be measured in inches, a Camden 



Fin. 3. — BRADPORD RAIN-GAUGE 



measure-glass is necessary for the estimation of small 

 amounts of rainfall. These glasses hold half an inch 

 and'are graduated in tenths and hundredths. There arc 

 measure-glasses also graduated in millimetres and tenths. 

 They hoM 10 millimetres (approximately 5 inch). 



