REPORT OF THE SECRETARY. 23 



« 



analysis of the recurrence of cold periods during the winter; tables 



deduced from the daily observations for ascertaining the corrections 



required to be applied to observations made only once or twice a 



day in order to obtain the mean temperature of places within the 



arctic circle; and, finally, observations to determine the diminution 



of temperature with an increase of elevation. 



Beside the corrected records of the motion of the air, the second 

 part of the memoir contains the resultant direction, the average 

 force, the mean velocity, the quantity, the frequency, and the dura- 

 tion of the winds. The third part contains not only the record of 

 the pressure, but also a comparison of the mercurial and aneroid 

 barometers, the diurnal and annual variation, the regular fluctuations 

 of the monthly and annual extremes of pressure. 



The expedition was supplied with thirty-six mercurial thermome- 

 ters, four maximum and minimum thermometers, twenty-four spirit 

 thermometers of different sizes, including two standards and a regis- 

 ter thermometer of thirty-six inches in length. A laborious series of 

 the different readings of these instruments, particularly at low tem- 

 peratures, was made, from which have been deduced corrections to> 

 be applied to the records prepared for publication. The differences 

 exhibited by the spirit thermometers at low temperatures was referred 

 to the unequal contraction of colored alcohol not chemically pure. 

 This liquid, when exposed to a great degree of cold, appeared to 

 change its condition, the coloring matter being deposited on the sides 

 of the tube. The lowest temperature observed during the first 

 winter, 1853-54, was, February 6, — 66°. 4; and during the second 

 winter, 1854- 55, occurred, January 8, — 65 .5. 



The highest temperature observed was July 23, 1854, +51°, giving 

 an absolute difference of 117°. 4. The diurnal maximum or highest 

 temperature of the day occurred in October and November, about 

 one hour before noon, and in April and May, three hours after noon. 



In the months of October, November, and December there are two 

 points of low temperature each day — one at 6 a. m., and the other at 

 about 9 p. m. ; during the remaining months of the year there is one 

 minimum during the twenty-four hours, which occurs at 1 a. m. 



It was a fortunate circumstance that the observations extended 

 over two winters, and thus gave a more exact mean for that season. 

 The warmest month is July, the coldest March; the temperature of 

 December, however, does not differ much from the latter. The 



