corresponds with the terrestrial pole than does the thermal 

 equator, which connects together the hottest points of al] 

 meridians with the geographical equator. Arago concludes, 

 from the gradual decrease of mean temperatures, that the 

 degree of cold at the northern terrestrial pole is 13, if the 

 maximum cold ohserved by Captain Back at Fort Reliance 

 (62 46' lat.) in January, 1834, were actually 70 ( 56-6 

 Cent., or 45 0> 3 Reaum.).* The lowest temperjtlure that, 

 as far as we know, has as yet been observed on the earth, is 

 probably that noted by Neveroff, at Jakutsk (62 2' lat.), 

 on the 21st of January, 1838. The instruments used in 

 this observation were compared with his own by Middendorff, 

 whose operations were always conducted with extreme ex- 

 actitude. Neveroff found the temperature on the day above 

 named to be 76 (or 48 Reaum.). 



Among the many grounds of uncertainty in obtaining a 

 numerical result for the thermal condition of the regions of 

 space, must be reckoned that of our inability at present to 

 ascertain the mean of the temperatures of the poles of great- 

 est cold of the two hemispheres, owing to our insufficient ac- 

 quaintance with the meteorology of the antarctic pole, from 

 which the mean annual temperature must be determined. I 

 attach but little physical probability to the hypothesis of Pois- 

 son, that the different regions of space must have a very va- 

 rious temperature, owing to the unequal distribution of heat- 

 radiating stars, and that the earth, during its motion with the 



* Arago, Sur la Temperature du P6le et des espaces Celestes, in the 

 Annuaire du Bureau des Long, pour 1825, p. 189, et pour 1834, p. 192; 

 also Saigey, Physique du Globe, 1832, p. 60-76. Swanberg found, from 

 considerations on refraction, that the temperature of the regions of space 

 was 58.5. Berzelius, Jahresbericht fur 1830, s. 54. Arago, from 

 polar observations, fixed it at 70 ; and Pectet at 76. Saigey, by 

 calculating the decrease of heat in the atmosphere, from 367 observa- 

 tions made by myself in the chain of the Andes and in Mexico, found it 

 85 ; and from thermometrical measurements made at Mont Blanc, 

 and during the aeronautic ascent of Gay-Lussac, 107-2. Sir John 

 Herschel (Edinburgh Review, vol. 87, 1848, p. 223) gives it at 132. 

 We feel considerable surprise, and have our faith in the correctness of 

 the methods hitherto adopted somewhat shaken, when we find that 

 Poisson, notwithstanding that the mean temperature of Melville Island 

 (74 47' N. lat.) is 1 66', gives the mean temperature of the regions 

 of space at only 8'6, having obtained his data from purely theoretical 

 premises, according to which the regions of space are warmer than the 

 outer limits of the atmosphere (see the work already referred to, $ 227, 

 p. 520) ; while Pouillet states it, from actinometric experiments, to be 

 as low as 223-6. See Comptet Rendus de I' Academic det Science!, 

 torn, vii., 1838, p. 25-65. 



