G02 PROCEEDINGS OP THE AMERICAN ACADEMY. 



maximum frequency of clear days about the time of the last quarter of 

 the moon, corresponding with the time of maximum pressure at Paris, 

 Carlsruhe, and Strassburg. The curve for Krakau, however, shows the 

 opposite condition. For purposes of comparison Van Bebber collected 

 in tables the results of various investigators. Tlie results showing the 

 relation of the moon's phases to atmospheric pressure ai'e given in Table 

 I., except that two stations where the interval of observation did not 

 exceed five years are omitted. In a large proportion of these results 

 the minimum of barometric pressure is about the time of the moon's 

 second octant and a maximum about the time of the last quarter. The 

 range from maximum to minimum is small, but the stations cover a con- 

 siderable portion of Europe and there is one as distant as Batavia. The 

 general agreement of tlie results is surprising in view of tlie fact that 

 some of the investigators were seeking to disprove the lunar period in 

 atmospheric phenomena, or at any rate were sceptical of its existence. 

 It is improbable that the pressure is high over all the earth at the same 

 time, so tliat fnture investigators will probably find opposite phases of 

 the period for different parts of the world. 



In Plate II. are curves plotted from the mean temperature for each 

 day of the lunar synodic period. They are drawn by means of a panto- 

 graph from curves given by J. Park Harrison in the Proceedings of the 

 Royal Society of May 4, I8G0, except No. IV., which I derived from 

 thirteen years' observations at the Blue Hill Meteorological Observatory 

 (1886-98). The curves were plotted from the unsmoothed mean of 

 each day of the period. No. I. is from the daily mean of temperature 

 at Greenwich, Eng., from 1856-G4. No. II. is from the minimum tem- 

 peratures for the same interval at the same [ilace. No. III. is from the 

 daily means of temperature at Oxford, Eng., from 18o9-Gl. No. IV. is 

 from the daily means of temperature at Blue Ilill, Mass., from 1860-98. 

 No. V. is from the daily means of temperature at Oust Silosk, Siberia, 

 from 1837-43. The maximum and minimum values are given by 

 numerals printed near the curves. The curves are irregular and the 

 rano-es not large, but they all agree in showing a generally higher tem- 

 perature between new moon and full than between full moon and new. 

 At the European stations the warmest weather seems to be about the 

 first quarter, and the coldest about the last, thus agreeing approximately, 

 either directly or inversely, with the times of maximum found by other 

 investigators for the pressure and for the number of rainy days. 



In Plate III. are curves showing the number of thunderstorms on 

 each day of the moon's synodic period. No. I. is plotted from results 



