THE METEOKOLOGY OF EDINBURGH. 65 



mean of Edinburgh, on the assumption that the distribution of pressure over the country 

 was normal. These corrections were obtained from Dr Buchan's paper on " The 

 Mean Atmospheric Pressure of the British Isles." * Although but little weight was 

 attached to the values thus corrected, they were of much interest when viewed in 

 connection with anomalies in the barometric gradients over the country. Maps were 

 prepared for a period of thirty -seven years, viz., from 1781 to 1817. It was not 

 necessary to adopt this tedious process after 1817, as from that date the instruments 

 were on the Fortin principle, and carefully observed. From an examination of the 

 results thus graphically shown by the data delineated on the maps, the elimination 

 of errors was rendered comparatively easy. I believe that the means thus obtained 

 give a close approximation to the average pressure for the period under discussion. 

 The observations utilised from 1817 to 1856 were the following: — From 1817 to 1826 

 the means were computed from the Calton Hill Observatory, where daily readings were 

 taken at 8 a.m. and 10 p.m. These were printed monthly in extenso in the Scots 

 Magazine for the years to which they refer. Adie's observations given in the Edin- 

 burgh Journal of Science were adopted for the period 1827 to 1832, while the Royal 

 Society's observations were employed from January 1833 to October 1834, and again 

 from 1839 to 1852, the hiatus being filled in from a register kept at Lasswade, six 

 miles S.E. of Edinburgh. The Lasswade means were calculated from 1828 to 1843, 

 so as to allow of the determination of the instrumental correction by comparison 

 with Edinburgh. Means were also computed for part of this period from the Dollar 

 register, which furnished an additional check. The hours of observation were, at 

 Lasswade, 8 a.m. and 10 p.m., and at Dollar, 9.15 a.m. and 8.30 p.m. The observations 

 at the rooms of the Royal Society from 1839 to 1852 were taken at 10 a.m., and were 

 deficient on Sundays and holidays. It was, therefore, necessary to interpolate values 

 for the missing days. The height of the barometer for these days was found from the 

 contemporaneous registers kept by Alex. Adie till 1850 and continued for some years 

 thereafter at his place of business. As the Royal Society observations were made only 

 once a day, it was necessary to reduce Adie's 10 a.m. and 10 p.m. readings, in order to 

 obtain corrections to be applied so as to bring the former series to the mean of 10 am. 

 and 10 p.m. This was accordingly done. The reason Adie's observations were not 

 utilised for the actual means is that there was no attached thermometer. The readings 

 could not, therefore, be reduced to 32°. The means for 1853 to 1856 were obtained 

 from Sir Henry James't abstracts taken in Edinburgh by the Royal Engineers. From 

 1856 down to the present time the 9 a.m. and 9 p.m. observations made at the Edin- 

 burgh stations of the Scottish Meteorological Society have, as already stated, been 

 employed. Every effort has been made to make the results comparable by reducing or 

 otherwise correcting the means to those of 9 A.M. and 9 p.m. For many years the hours 

 were 8 a.m. and 10 p.m., or 10 a.m. and 10 p.m. ; observations taken at these hours differ 



* Jour. Met. Soc, vol. vi. pp. 14-18. 

 t Abstracts from Meteorological Observations taken at the stations of the Royal Engineers. 



