METEOROLOGY. 295 



otor, for it is eerfaiiily more scientific; aiirl importnnt that we slioiiM in 

 tliis way obtain a correct measure of the gradient force that produces 

 the movements of the atmosphere, and it is much to be desired that 

 this reduction shouhl be applied generally to individual observations 

 rather than to monthly and annual averages. As the Signal Service 

 has announced the adoption (on and after January 1, 1885) of this cor- 

 rection in its publications, he hopes that the co-operation of other me- 

 teorologists may be prompt and general. (7>. M. Z., i, p. 3l';3.) 



143. Prof. R. Bornstein, of Berlin, has discussed the value of the ingen- 

 ious suggestion made by Professor Nipher, of Saint Louis, that a protec- 

 tion in the form of a funnel be given to a raiu-gauge in order that the 

 influence of eddies of wind around the mouth of the gauge may be 

 diniini.shed. After recounting some of the theories explanatory of the 

 diminution in the catch of rainfall, all of which have been reduced 

 to one, namely, the greater formation of eddies by the stronger winds 

 at high exposures, he states that Nipher's proiJOsi'tiou seeming per- 

 fectly rational it was tested by experiments by himself. Two simi- 

 lar rain-guages, having receiving surfaces of 500 square centimeters, 

 such as are used at the stations of the Deutsche Seewarte, were estab- 

 lished near each other, one protected, the other unprotected ; both were 

 on the roof of the Berlin Agricultural High School, near the southwest 

 corner, and 20™ above the ground. The jiosition of the gauges was 

 changed several times within narrow limits on the roof. The greatest 

 individual difference between the catches of the two gauges was during 

 alight rain in January, and amounting to over 500 per cent., the pro- 

 tected gauge giving the larger figure. The annual average difference 

 was about 11.5 per cent. There is therefore no doubt as to the general 

 value of Nipher's invention ; but this becomes still more remarkable if 

 we classify the precipitation, whence we find the ratios or the catch of 

 the unsheltered divided by that of the sheltered gauge. The figures are 

 as follows : For snow, the ratio is 0.54 ; for fine rain, 0.66; for rain and 

 snow, 0.82 ; for rain with hail, 0.89 ; for rain, 0.90; for heavy rain, 0.9!J; 

 whence it seems that Isi[)her's device gives the rain-gauge a better pro- 

 tection, and that it therefore makes better measurements in proportion 

 as the wind can more easily deflect the precipitation ; and in general his 

 device does better in i)roportion as the wind is stronger. Perhaps the 

 most important defect in the Nipher gauge is that with heavy snowftiU 

 the protecting funnel is soon filled up, and then ceases to work. (/>. M- 

 Z., I, p. 381). 



144. Dr. C. Lang, of Munich, has experimentally investigated the <pian- 

 tity of rainfall as dependent on the shape of tlierain-gauge, an investiga- 

 tion which, as he says, is the more important now that weare establishing 

 such a great additional number of rainfall stations and are about to 

 discuss the climatic conditions of various portions of limited regions, 

 where the natural differences will be slight and comparable with the 

 errors of observation. In general we do not in meteorology expect ab- 

 solutely correct measures of auy element, but are satisfied if uniformity 



