93 
of January being 2 *5° Cent. (4*5° F.), and March 11th to 
20th 2 9° Cent. (5*2° F.) warmer than mine, it would seem 
that the difference principally arises from different screening 
and placing of the instruments. The screen-box is so placed 
in a corner that it will receive reflected heat from two walls 
and probably the readings are too high on that account. 
As I pointed out last year, the question of screening is 
most difficult in such a climate, and therefore I was very 
glad towards the end of the season to accept from Dr. Berry 
the loan of one of the Swiss metal cylinder-screens, which 
was 50 centim. high and 25 centim. in diameter. 
These are usually placed just outside an upper window 
and the figures to which I have referred were so taken. I 
however placed it with a board below, as sometimes used by 
the Swiss, and put this on a stand near to my own screen. 
I then placed the lid of a large box about half a metre 
away in a sloping position to the south of the thermometer 
cylinder. This formed a shading root and roughly made 
a protection similar to that recommended by Professor Wild. 
The metal screen was entirely in the shade from 9.30 and 
but little sun shone on it before that, so that the one o’clock 
observation would be uninfluenced by the sun. The next 
question is at what height should the thermometer be 
placed. I take it that one of the objects of the screen 
is to prevent terrestial radiation from the thermometer, and 
that therefore the thermometer should be well protected 
by the cylinder, but having seen the thermometer at the 
lower level of the cylinder in one of the Swiss stations, 
I placed one thermometer about one inch below the level of 
the cylinder so that it would be always kept in shade by it 
but be able to radiate out heat ; another one I placed near 
the middle, so that it would receive any heat from the metal 
if it was warmer than the air. 
The upper thermometer which, as explained, was fully pro- 
