82 Proceedings of Eoyal Society of Edinburgh, [jan. 17, 
would be taken to keep the polish in as perfect a condition as 
possible. 
In addition, however, to these possible sources of error, there is 
in the situation of the two places where the observations were made 
an essential difference which would affect the results. The site of 
the screens at Granton was very freely exposed to every breath of 
wind, being on a knoll in the middle of a field near the sea shore, 
perfectly open to the west, north, and east, while the land rose a 
little to the south, and the screens were at a great distance from 
trees or anything that could obstruct the free circulation of the air. 
The site on which the screens are placed here is very different, and 
not so good in many respects, though I think it will compare favour- 
ably with the site of many screens in daily use. Here the screens 
are on a lawn, and surrounded at no very great distance by trees. 
As the surroundings of the screens seem to be a matter of greater 
importance than might at first be thought, it will be as well that 
I state more fully the conditions surrounding the screens here. 
Standing at the screens, the view in the different directions is closed 
in principally by trees. The view to the south and round by west 
to north-west is closed in by a narrow line of trees running north 
and south, at a distance of 26 yards at the point where it comes 
nearest to the screens. From the north-west to the north-east, at a 
distance of about 24 yards, there is a holly hedge 8 feet high, and 
also a few trees. From north-east to east are trees at a distance. 
From east to south-east a holly hedge running north and south, and 
coming at its north end to within 12 yards of the screens. From 
south-east to south are stables and kennels at a distance of 35 yards. 
The ground slopes slightly down to the north. The screens are 
placed east and west of each other, and all the ground in view is 
under grass. It will be seen that, while the position is somewhat 
sheltered from winds blowing from east to south-east, it is fairly 
open to winds from south to south-west, as well as from north-west 
and north-east. But, on the whole, it is evident the site is much, 
more confined than the Granton one. 
An evident result of the opener exposure of the Granton site is, 
that the Stevenson screens would be kept cooler there, on account of 
the freer circulation of the air through the louvres, the screens 
would thus be both more nearly correct, and therefore nearly agree 
