If we reckon in this way, and take into account the horizontal refraction and the fact that the obliquity of the ecliptic about the year 500 was approximately 11' greater than now, we shall find that at that time the midnight sun was seen for forty days (i.e., from June 2 to July 12) in about 66° 54' N. lat., or in the neighbourhood of Kunna, south of Bodö; but at the same place more than half the sun's disc would be above the horizon at noon at the winter solstice; it was therefore not hidden for a single day, much less for forty days. But, on the other hand, it was not until 68° 51' N. lat., or about Harstad on Hinnö, that they had an unbroken winter night, without seeing the rim of the sun, for forty days (from December 2 to January 11); but there they had the midnight sun in summer for about sixty-three days. The fable of a summer day of the same length as the unbroken winter night cannot therefore have originated with the Northerners; it must have been evolved in an entirely theoretical way by astronomical speculations (in ignorance of refraction) which were a survival of Greek science, where the length of the northern summer day was always assumed to be equal to that of the winter night. But that information had been received at this time from the Northerners is probable, since the statement of a forty days' summer day and winter night is not found in any known author of earlier date,[133] and Jordanes' contemporary, Procopius, has an even more detailed statement, especially of this winter night (see later). The probability is that what the Northerners took particular notice of was the long night, during which, as Procopius also relates, they kept an accurate account of the days during which they had to do without the light of the sun, a time in which "they were very depressed, since they could not hold intercourse." This must also have been what they told to the Southerners, while they did not pay so much attention to the length of the summer day, when of course they would in any case have plenty of sunlight. We must therefore suppose that the latitude worked out according to the winter night of forty days is the correct one, and this gives us precisely Sophus Bugge's And--Andö, or, better still, Hinnö.