68 Meteorological Influence of the Moon. 



and the extremes of heat and cold occur less frequently be- 

 tween the new moon and the first quarter, than during the 

 other parts of the period. But a much longer series of careful 

 observations is required to verify plienomena of this kind, and 

 to establish the law which they observe. 



In investigating this subject, remarks M. Madler, I was 

 soon convinced that those great barometrical oscillations whose 

 causes are still unknown, as well as the anomalies of the state 

 of our atmosphere, must so inevitably mask the trifling influ- 

 ences of the moon, that we must needs for a long time re- 

 nounce the hope of obtaining any permanent results from the 

 observations made in our higher latitudes. The extent of the 

 barometrical variations, according to the preceding observa- 

 tions, is 26 lines at Berlin (comprised between 321 and 347 

 lines, at an elevation of 130 French feet above the ocean) ; and 

 those of the thermometer in the shade amount to 53° R. (viz. 

 from — 23° to -|- 30°) ; there are some whole years in which 

 the diurnal period of the barometrical height is nearly entirely 

 disfigured : since a single considerable oscillation, and we 

 have witnessed 14 lines in twelve hours, may notably change 

 the form of the annual curve. These immense inequalities 

 disappear near the tropics, and there the extreme oscillations 

 are reduced to two or three lines, and we can every day recog- 

 nise the solar period. I have endeavoured, therefore, to pro- 

 cure good observations made in the neighbourhood of the 

 equator, — and I have obtained, by the kindness of States- 

 Councillor Schumacher, a complete copy of those made five 

 times a-day, from the 20th February 1829, to the 31st of Ja- 

 nuary 1833, by Messrs Trentepohl and Chenon, at Christians- 

 burg in Guinea, at the latitude of b^° N. The barometrical 

 range is so constant in this place, that each observation, after 

 the usual reductions, and the necessary corrections from the 

 known periodical variations have been made, scarcely departs, 

 except in a very small number of instances, even a single line 

 from the general annual mean. 



M. Madler, after reporting the monthly means of these ob- 

 servations for every hour when they were made, draws the 

 conclusion, that in this locality two periods of barometrical 



