Observaticms of the United States Signal Service. 5 



We see from this table that the stations at which the ther- 

 mometer rose most above the mean were all north of Int. 39°. 

 and they are all situated between the Rockv Mountains and the 

 meridian of 81° westof Greenwich. We also see that a high tem- 

 perature was generally accompanied bv a barometer below the 

 mean; the average pressure shown in column fourth being 29*70. 

 Column fifth shows that on each day there was a considerable 

 barometric depression in some part of the United States and the 

 center of this depression was always north of lat. 40°, and the 

 region of greatest heat was always within the influence' of the 

 area of depressed barometer. In four instances the station 

 which showed the greatest rise of temperature showed also the 

 greatest depression of the barometer. 



In order to exhibit more clearly the form and position of 

 these heated areas I have prepared ("hart I, which shows the 

 thermic isabnormals for June 18th, at 4 h 35 ra , P. it., the curves 

 being drawn at intervals of 5°. 



From this chart we perceive that near the Gulf of Mexico. 

 and also near the northeast and northwest angles of the United 

 States, the temperature was below the mean for the month of 

 June, but in every other portion of the United States the tem- 

 perature was above the mean for the month. In order to de- 

 termine how much the mean temperature at the hour of obser- 

 vation exceeds the mean temperature of the month, I have 

 compared the observations of maxima and minima tempera- 

 tures at all the Signal Service stations as far as they have been 

 published in the annual Eeports. I have taken the mean of 

 the maxima of temperature for the month of June at each 

 station, and also the mean of the minima. The difference be- 

 tween these two months I call the mean diurnal oscillation of 

 the thermometer. I ha arison for each of the 



years 1873, 4, and 5. The following table shows all the Signal 

 Service stations at which the mean diurnal oscillation exceeds 

 22°, and also all those stations at which the oscillation is less 

 than 15°. Column second shows the mean diurnal oscillation 

 for the month of June at the stations named in column first; 

 column third shows the number of years of observation; col- 

 umn fourth shows the height of the station above the sea ; 

 column fifth shows the annual rain-fall at each station; and 

 column sixth shows the rain-fall in 1873, from June 12th-26th 

 inclusive. 



Since we have a series of hourly observations of the ther- 

 mometer at Philadelphia and Toronto we can compute the cor- 

 rection to be applied on account of the hour of observation at 

 each of the preceding stations, if we assume that the i 

 at different stations is proportioned to the entire daily oscilla- 

 tion. We thus find that at the hour of 4 -35 P. M., Washington 



