TTNCERTAINTY OF MODERN METEOROLOGY. 23 



have been deduced with respect to the hygrometrical and ther- 

 raometrical conditions, to the climate, in short, of different coun- 

 tries, have very often been derived from observations at smgle 

 points in cities or districts separated by considerable distances. 

 The tendency of errors and accidents to balance each other au- 

 thorizes us, indeed, to entertain greater confidence than we could 

 otherwise feel in the conclusions drawn from such tables ; but it 

 is in the highest degree probable that they would be much modi- 

 fied by more numerous series of observations at different stations 

 within narrow limits.* 



forty-six inches at Cluny. — Nature, September 24, 1874. This enormous dif- 

 ference in measurement is too great to be explained by possible errors of ob- 

 servation or other accidental circumstances, and we must suppose that there 

 are, in different parts of this small island, great differences in the actual pre- 

 cipitation, but still much of this variation must be due to causes whose range 

 of influence is extremely limited 



In 1859 at Charleston, S. C, there fell at the U. S. Arsenal, in a single rain- 

 storm of an hour and a half, two inches of water ; at the Register's Office in 

 the same city, at the distance of two miles from the Arsenal, the fall was but 

 one-third. of an inch. 



In the same year observations at three stations in the city of San Francisco, 

 gave a total rain-faU of 16" 34 ; 25" 41 ; and 21" 39, respectively, and during 

 the whole period from 1853 to 1860, the meteorological records at different 

 stations in the same city show similar discrepancies. See Smithsonian Contri- 

 butions, Vol. XVIII., p. 143. 



Like differences are constantly found in the temperature registered at differ- 

 ent stations in the same vicinity, and it is obvious that in such cases no trust- 

 worthy conclusions as to the general meteorology of territories of even very 

 small extent, can be obtained by mere averages of interpolations. 



As every astronomical observer has his personal, so every meteorological 

 station has its local, equation, and the determination of these equations ought 

 to be a cardinal object in every system of pluviometrical or thermometrical 

 observations. Records of observations at the same hours and by the same 

 methods, with the same or other carefiiUy compared instruments, for a series 

 of years, may authorize conclusions as to the essential or accidental climatic 

 conditions of that precise locality ; but in the present state of our knowledge, 

 such records alone warrant no inference as to the meteorology of any other 

 point, even within the distance of a mile, unless it may be in the case of 

 observations with instruments absolutely insulated on great plains, or other- 

 wise placed in exactly corresponding positions. Hence, until the equations 

 we speak of are ascertained, results deduced from the comparison of observa- 

 tions made at different periods and at different stations can have no scientific 

 value whatever. 



* The nomenclature of meteorology is vague and sometimes equivocal. Not 

 long since, it was suspected that the observers reporting to a scientific institu- 



