2 RECORD AND DISCUSSION OF TEMPERATURES. 



again by Mr. Sonntag, by means of Mr. Tagliabue's standard. Besides these, there 

 were four ma.ximum and four minimum thermometers, and two dozen spirit ther- 

 mometers of various sizes, including two standards with a register 36 inches in 

 length.^ By one of these, and a mercurial standard of the same length, most of 

 the temperatures of the air were noted. 



Dr. Kane remarks :^ " The temperature on the floes was always somewhat higher 

 than at the island, the difference being due, as I suppose, to the heat conducted by 

 the sea-water, which was at a temperature of +29°, the suspended instruments 

 being affected by radiation." 



This was on the 17th of January, 1854. On another page, he says:^ "Upon' the 

 ice-floes, commencing with a surface temperature of — 30°, I found, at two feet deep, 

 a temperature of — 8°, and at four feet +2°, and at eight feet 4-26°." This was 

 in midwinter, on the largest floe in the open way off Cape Stafford. This subject 

 will again be referred to. 



Comparison of Thermometers. — The different readings of the instruments, par- 

 ticularly at temperatures below — 40°, made their frequent comparison, in order to 

 obtain corrective elements, a matter of great importance. Appendix No. XI. of 

 the second volume of the narrative, contains a full exposition of the unreliable in- 

 dications of the instruments at very low temperatures, and to this appendix tlie 

 reader may be referred for further details.'' Whether these anomalies be due to 

 irregularities in the diameter of the tubes, or to unequal contraction of colored 

 fluids of different specific gravity, it is admissible to suppose that the errors for 

 a number of instruments, compared at the same temperature, may be as fre- 

 quently in excess as in defect, provided they keep within a certain limit, beyond 

 whicli the indications become useless. Tlie mean reading of all thermometers com- 

 pared at a certain temperature has, therefore, been taken for the true temperature, 

 and, by comparing each result with this mean, a series of connections has been 

 obtained for each instrument. The same view was taken by Dr. Kane. 



Fi'om the comparisons of February 5th, 6th, and 9th, 1854, I was led to suspect 

 that some or all of the spirit thermometers, designated in the original log-book Nos. 

 1, 2, 3, 4, 5, were probably identical with those in the table of Appendix No. XI. 

 in Vol. II. of the narrative, there named C, B, No. 4, A, No. 12, respectively. The 

 numerous comparisons given in that appendix, and to which a few more have been 

 added, made it unnecessary to use the observations from the above numbers in the 

 first set of comparisons between the temperatures — 68° and — 20°. In the follow- 

 ing table, arranged according to temperatures, S denotes the 36 inch spirit standard, 

 upon which instrument the temperatures determined by the expedition mainly 

 depend; aS'joj, a second similar standard; M, the 36 inch mercurial standard; the 

 rest are alcoholic thermometers, from twelve to eighteen inches in length of scale. 

 Fahrenheit's scale is used throughout. 



' The expedition was also provided with one or more ether tliermometers, of which I could find no 

 further account. 



^ See Narrative, Vol. I. p. 154. 



» Narrative, Vol. I. p. 2G7. 



* An extract from this appendix will be found at the end of this article. 



