RECORD AND DISCUSSION OF TEMPERATURES. 



For temperatures between 20 and +15, the following table of corrections 

 has been constructed. The third column contains the number of comparisons. 



In the absence of direct comparisons for temperatures of +16 and upwards, I 

 have adopted Mr. Sonntag s corrections, as found by comparing the daily means for 

 May, June, and July, in the log-book, with the corresponding means in Appendix 

 XII. of Vol. II. of the narrative. They have, however, all been diminished by 

 0.7, the correction, according to his table, applicable at +15.0, for which tempera 

 ture I have found that no correction was required. The last set of corrections to 

 the spirit standard becomes then 



In the following abstract of the hourly record of the atmospheric temperatures 

 from September 1st, 1853, to January 24th, 1855, observed at Van Rensselaer 

 Harbor, the corrected figures have been inserted in accordance with the previous in 

 vestigation; foot notes contain any additional information that may be required. 



These temperatures refer to the level of the sea. The observations were made 

 by the officers and men on duty, and are referred to mean local time. Occasional 

 short interpolations were effected by means of the known diurnal variation, and 

 the gradual change in the absolute temperature ; in all cases, accordingly, the 

 means given are corrected for any such omission in the record. The hourly series 

 in my possession terminates with January 24th, 1855; the daily means up to April 

 30th, 1855, have been extracted from the second volume of the narrative. Before 

 these numbers were set down, the difference in the system of corrections, as adopted 

 by Mr. Sonntag, and in the present paper was applied. 



That the temperature was lower in winter in exposed positions at the astro 

 nomical observatory, and in the outer bay or channel, than at the meteorological 

 observatory on the floe in the harbor, there can be no doubt ; but, owing to non- 



