Mkteorolocy. 31 



from seeing distinctly its outlines. I could see, however, that it 

 was a circular fort, whose rampart was mainly formed of stones, 

 and that its size was not larger than the camp I had just left. 



17th Jnmiarf/, 1896. 



The Rev. William Andson in the cliaii-. 



Ne7i' jMe>nbers.—^lQ^?<vs Oeorg-e Irving-, Xewcastle ; William 

 D. Mackenzie, Henley-on-Thames ; and Samuel Smith, M.P.. 

 Liverpool. 



Donations. — Report of the British Association for the 

 Advancement of Science. 



CoinruNic'ATioNS. 



I. — Report of the Meteorology of Dumfries for i8qj. 

 By the Rev. William Axdson. 



Barometei;. — The highest reading of the year occurred on 

 the ;31st January, when it rose to SO'SO.j in., and the lowest on 

 the 28th March, when it fell to 28-593 in., giving an annual range 

 of 2-212 in. There were other three occasions on which the 

 barometer fell below 29 in. — (he first in the middle of January, 

 the second between the 9th and 12th November, and the third in 

 the middle of December. Althoug-h the lowest reading registered 

 was 28-593 in. on the 28th March, there is reason to believe that 

 in the early morning of the 11th November the barometer touched 

 a still lower point. This may be inferred from the circumstance 

 that where readings were taken every hour dui'ing the night 

 between the 10th and the 11th, considerably lower readings were 

 recorded. At Leith, for example, where this was done, a reading 

 of 28-352 in. was registered at 3 a.m. The reading of 28-G07 in. 

 was registered here at 11 p.m. of the 10th, but the mercury at 

 that hour was still falling, and in all probability, if a reading- had 

 been taken two hours later, say at 1 a.m., or two hours before the 

 Leith observation of 28-325 in., it would have been equally low. 

 The mean pi-essure for the year (reduced to 32 deg. and sea level) 

 was 20-893 in., which is a little below the averag-e of th;^ last nine 



