356 Dr Pearson, Observations of the Sun on the [Mar. 8, 



northern meridian, its change in altitude is 15", 1' 0", 4' V\ 

 16' 2", respectively. Now in the latitude of the North Cape, 

 say 71° 11', with the Sun's declination IT*' 57' 50", as it must have 

 been nearly, the depression of its centre below the horizon must 

 have been about 51' 10". Allowing 15' 48" for its semidiameter, 

 and 4' 5" for the dip, this is reduced to 31' 17". According to 

 Ivory, the normal refraction for Z.D. 90" (I do not claim, as I 

 ought, for 90° 4' 5") is 33' 24", or augmented for the barometer at 

 30'25, about 34' 50"; in which case about 3|- minutes of the Sun's 

 upper limb ought to have been visible; now about 45 minutes 

 before northing its altitude (upper limb) was only 9' 50", and, what 

 is more, by 35 minutes before, had dropped to 2' 50", lingering I 

 own at what, as far as I could judge, was about an altitude of 1' 

 for some time, though from the reflected rays the true horizon of 

 the sea was extremely doubtful. It is almost impossible to give 

 any explanation which will account for the disk of the Sun being 

 so depressed, except by an actual change in the refraction of the 

 atmosphere. 



I will now mention something which leads me to think (for I 

 cannot build a stronger argument on a single example) that the re- 

 fraction for very small altitudes is extremely uncertain. On the 

 previous evening, about 12 o'clock, we were passing a fine and well- 

 defined headland named Nordkyn, about 30 miles E. by S. of the 

 North Cape, and it occurred to me to take the Sun's altitude, which 

 for the upper limb proved to be 16' 40", the time, 12h. 24m. 55s., 

 local apparent time, being accurately noted. I did not take more 

 than one sight ; but on working it out twice very carefully with the 

 help of the chart, I found the theoretical and observed altitudes of 

 the Sun to coincide within a very few seconds. It is certain that 

 the influence of the Gulf Stream both on the water of the sea and 

 on the atmosphere diminishes very rapidly to the east of the 

 North Cape ; whether so great a change in the refraction can be 

 attributed to a change in the state of the atmosphere, it is not for 

 me to decide; but I can say this, that no pains were spared by 

 myself to take my observations carefully. 



Yet, on the other hand, with a constant temperature and 

 barometer, the refraction seems almost certain. In the Monthly 

 Notices of the Royal Astronomical Society for the year 1863 (vol. 

 XXili. p. 58) and for 1864 (vol. xxiv. p. 26) there are given two 

 series of observations taken by Sir A. Lang on the disk of the Sun 

 when rising above the sea-horizon, at St Croix in the West Indies, 

 at an elevation of 440 feet above the sea level. The temperature 

 throughout was extremely uniform, only varying between 73" F. 

 and 79"5, with a mean of 77°'5 in the first set of 17 observations 

 (which extended over many years, from 1831 to 1848), and between 

 76" and 79"-3, with a mean of 76"-94; in the second set of 18, 



