MEtEOROLOGY. 



297 



4. A simple cylindrical ganjje 25.2 aperture, or ^^ square meter, 

 wbicb, of course, luivinj? no conical surface, should give results free from 

 any error the latter may introduce. 



This latter proved wholly uiMellableiu drifting snow ; the comparison, 

 therefore, between these as snow-gauges was given up for the i)resent, 

 but as regards rainfall during the summer mouths the following results 

 are expressed in percentage of the catch given by No. 1, namely : No. 1, 

 100 ; No. 2, 97.5; No. 3, 95 ; No. 4, 99. The small catch of No.'.'} may be 

 due partly to the generalpoorresultsgiven by small apertures, but if pre- 

 vious hypotheses were correct must also be partly due to the increased 

 ratio between the inclined conical surface and the aperture of the gauge. 

 These two results, however, suggest that so far as the heavier rainfall of 

 the year is concerned one need not be very anxious about the instruments, 

 provided thej- are not too small ; the more important question will be as 

 to the proper exposure of the gauges and radical differences in the struct- 

 ure or shapes of the gauges. At Miinich the gauges of the Central 

 Meteorological Station, and that at the Observatory of Boge^ihausen 

 2 7 kilometers apart, on the average of several years, with comparable 

 instruments of this same pattern, show that 8.G i)er cent, less precipita- 

 tion fell on the right than on the left bank of the river Isar, while for a 

 ])eriod of nine months of special comparative observations 16 per cent, 

 less fell on the right than on the left side. Special observations are now 

 being made in this region at three other stations, namely, on the roof 

 of one of the public buildings of the city and on the grounds of two of 

 the observatory buildings. The differences between the records are such 

 that at least tive stations would seem necessary if we would have a cor- 

 rect result as to the distribution of rain over a space of one English mile 

 square, or at the rate of 3,500 stations for one division of Bavaria. We 

 must conclude, then, that meteorology will jn'otit least of all by the ordi- 

 nary rainfall measurements, although for hydrography and engineering; 

 they are very important. (/). .1/. Z.^ i, p. 431.) 



145. G. J. Symons, in the British Rainfall for 1880, gives a collation' 

 of results as to the effect of elevation above the ground upon the quan- 

 tity of rain caught by the rain-gauge. The following table shows the- 

 percentage of the upper gauge as compared with the amount caught b^r 

 a gauge at the surface of the ground : 



