436 F. H. Bigelow — Inversion of Temperatures in 



The meteorological observations of the U. S. Weather Bureau 

 are now made at 8 A. M. and 8 P. M. 75th meridian mean time, 

 and as it is generally agreed that the arithmetical mean of 

 these values is nearly equal to that derived from hourly obser- 

 vations, this has been used in reducing the variations from day 

 to day. In some of the years, if necessary, the mean of the 

 maximum and minimum readings is taken as the mean for the 

 day. In every case this mean corresponds to a noon observa- 

 tion at the station, as with European stations, and therefore a 

 correction for longitude is required for the epoch of the ephem- 

 eris. I have preferred to retain the Greenwich epoch 

 throughout the computations, and then move the comparison 

 magnetic curve to the right by the amount of such correction. 

 It has been found that the variations of the meteorological 

 elements synchronize closely with the fluctuations of the mag- 

 netic curve in the extreme northwestern states, Manitoba, 

 North and South Dakota, and Montana. If the temperature 

 changes are transferred to millimeter paper, so that one centi- 

 meter is equivalent to a day, then for this region there is only 

 a small correction to the epoch of the corresponding magnetic 

 curve. If we could have our observations at a proper origin 

 in British Columbia, no correction would be needed, and such 

 an origin is 115° W. longitude, 55° N. latitude. This seems 

 to be physically determined by the peculiar meteorological con- 

 ditions of North America, namely, the latitude of the high 

 pressure belt, and the winter accession to the same from the 

 Arctic regions, which passes near this origin ; the neighborhood 

 of the pole of maximum cold and maximum magnetic intensity ; 

 the nearness to the magnetic pole in Boothea Island ; all of which 

 in combination cause this region to be a special theatre of for- 

 mation of the high and low areas that are the agencies causing 

 many of the weather variations of the United States. Passing 

 to the south and east of this origin one can follow the changes 

 produced by the temperature waves, by simply moving the 

 magnetic curve to the right a number of days equal to the 

 time of progress from the origin eastward. In practice, the 

 calendar dates have been written along the top of the sheet and 

 underneath them the dates of the magnetic ephemeris. Begin- 

 ning at the day with its fraction which starts a new period, a 

 line is drawn down the sheet so that it passes through the same 

 phase for each district of the country. This was always deter- 

 mined by matching the magnetic curve to the temperature 

 curve. Hence the same phases may be collected together by 

 counting this line as the beginning of a period in every case. 

 The difference between this line and the straight lines drawn 

 from the ephemeris date gives the time consumed in the east- 

 ward drift, Since in this way the longitude from Greenwich 



