358 Dr Pearson, Observations of the Sun, etc. [Mar. 8, 



I was able to get five sets of observations of the Sun on the 

 northern meridian with a six-inch theodolite graduated to 20", 

 used as an altazimuth, and also one good set with my prism-circle 

 on the sea-horizon, at Vardo or Wardhuis, in lat. 70° 22' 30". The 

 last set of 12 sights gave the Sun's altitude a little (1' 6") too 

 high, with tolerable uniformity : the Sun's altitude (centre) being 

 as nearly as possible 4°. With the theodolite, on the first night, 

 with the Sun's centre about 3" 45', its place came out again about 

 1' 0" too high; on the next, with the Sun's upper limb (the only 

 one observed) about 3", the error was 37" too high. The third 

 set (Sun's centre about 2° 20') gave, on account of clouds, a good 

 mean result, but with great discrepancies. The next evening 

 (Sun's centre 2° 15') gave for 12 observations a mean result of 37" 

 too low. The last set (Sun's centre 2° 0') gave an error of 27" too 

 high on the upper limb, and 25" too low on the lower limb. The 

 observations with the theodolite, it may be added, gave uniformly 

 the position of the Sun lower on the lower limb than on the upper 

 one, and as this could not well be due to the instrument, the position 

 of the Sun being marked by taking its disk when in contact with 

 the point where the cross wires intersect on the horizontal wire, it 

 seems possible to suppose that the refraction on either limb of the 

 Sun was not the refraction actually due to that required by the 

 altitude of that limb, but rather that due to some intermediate 

 point nearer the Sun's centre, which will account more or less for 

 the observed error. My reason for referring to these observations is 

 because, depending as they do on the adjustment of the instrument 

 by a spirit-level, and not on the sea-horizon, one element of uncer- 

 tainty is eliminated ; and secondly because they prove, if proof is 

 needed, that it is only within the narrow limit which I have defined 

 already as a degree and a half or two degrees, that the accepted 

 laws of refraction seem to become seriously uncertain in their 

 application. 



The problem of horizontal refraction is one which is far best 

 subjected to practical observation in very high latitudes, on account 

 of the nearly horizontal motion of the objects observed when near 

 the meridian, as I have already pointed out, pp. 355, 6. Had I been 

 able to find a suitable station to the West instead of to the East of 

 the North Cape, I have reason to think that the sky at night would 

 have been clearer than I found it where I stationed myself, and 

 that I might have been rewarded with a better series of observa- 

 tions with my theodolite of the Sun's motion in altitude and 

 azimuth simultaneously, which was the main object which I had 

 in view when I went, and which I regret to say was to a great 

 extent a failure, owing to the sky being almost always obscured 

 near the horizon when the Sun was at its lowest. 



