Sir,

While on a visit to Norway this summer, I saw a very striking example of the effect of atmospheric refraction. Quite early one morning, in brilliantly clear weather, I set up a theodolite in order to examine it, and amused myself by looking at surrounding mountain peaks through the telescope. I directed the telescope on to the summit of Galdhöppigen (the highest point in Norway), then looked at some other part of the instrument, and on looking again through the telescope I noticed that the instrument had apparently moved. Further examination disclosed that the supposed movement was the effect of change of refraction on the apparent elevation of the top of the mountain. The change was so rapid that the object seemed to move somewhat faster than the.apparent movement of the Sun due to the Earth’s rotation.

I was at Juvass, at an altitude of 1840 metres, on the shore of the small lake there. The line of sight passed over about three-quarters of a kilometre of ice-cold water, then for a kilometre or so over a sidelong slope of snow and rock facing east-south-east. The minimum height of the line above this surface was about 20 metres. The rest of the line was well above ground all the way, to a total Iength of about five kilometres.

The phenomenon described must have been caused by an exceptional combination of conditions in the presence of snow, rock and clear weather, but it points to the necessity for precaution in any work involving the observation of vertical angles. Observations should be arranged so as to minimize the effects of any changes of refraction.