KXI'KIHTION TO POINT P.AUROW, ALASKA. 



511 



onliiuiry hourly readings. During lss_>, one lamp was continually burning in the observatory, 

 but early next year three lumps wen-, kept burning, tlic supply of oil in StM8 being greater Ilian was 

 at iirt-t supposed. The annual average, temperature in tlic observatory, as shown by a Fahrenheit 

 thermometer inside the x.inc cover of the bililar, was + 1!.0 or 7.22 C. 



In consequence of the irregularities in the state of the instrument as shown by the monthly 

 mean readings, the only available method for deducing the temperature nellicient <j appeared to 

 be that of selecting a number of consecutive and umlistnriicd days at times when the temperature 

 was lapidly changing, and lind for each case the apparent, change of the daily means in scale 

 divisions corresponding to a change of 1 iu temperature. The following values were thus found: 



It is proposed to adopt provisionally the mean value 2.2 0.8 which is equivalent to a. 

 decri'itise of 0.000105 part of the horizontal force for an increase of temperature of 1 Fah. or 

 3 = 0.000165 



In the following table the values in columns 3, 4, 5 are unconnected for changes of temper- 

 ature, the next three columns show the temperature differences for which corrections were required, 

 and the last three columns give the diurnal variations thus corrected. The values are laid down on 

 the accompanying diagram. 



Solar-ditiriKtl curiation of the horhoii-tnl force, inclusive ofdistitr''finces, and expressed in parts of the 



force, at Ugla&mie, 1882-'83. 



At Uglaamie the daily maximum value of the horizontal force occurs between the hours 5 

 and 7 p. m., and the daily minimum about 7 a. in.; there is also a very slight indication of a 

 secondary disturbance in the regular progression between 3 and 5 a. m. corresponding to a second- 



