146 



Loiv hut Unreliable Thermometer Headings. [February, isbs. 



had been driven off shore. The mean of the five thermometers at 7 p. 

 m., when the gale had entirely subsided, was — 39°; but Hall had now 

 further reason to place no confidence in two out of seven instruments, 

 one of which stood at — 100° and the other at — 75°. Shortly after 

 this he wrote : '' It is annoying to have but one of nine thermom- 

 eters, right. But by taking even one to the United States, and having 

 it compared with a standard, with my data, all the observations can be 

 worked up to said standard." 



During the night of the 26tli his five self-registering thermometers 

 read, —48°, —46°, -48°, —48°, — /)2°. At 8 a. m. he experi- 

 mented with the mercury given to him by Mr. Green, one of the in- 

 strument-makers of New York, for his artificial horizon. Pouring some 

 of this into a dish near his thermometer, he found the mass quickly 

 frozen, small spherical drops remaining fluid until the pressure of a 

 pencil changed their form. When the mass of the mercury again 

 became fluid, or nearly so, with the rising temperature, these globules 

 remained solid. 



The following table gives the results of his observations ; the 

 thermometers numbered I, II, and VII, at first being below the marks, 

 were not read : 



