£ 2 6 Sir george shuckburgh’s Obfervations 
until I returned; the. windows and doors of the room, 
in which the inftrumerit was placed, being left open, by 
which means there was a free communication with the 
outward air, and the barometer not expofed to the Sun. 
The detached thermometer was hung on the window 
towards the' north-eaft, where there was neither direcSt 
nor reflected- heat from the Sun n> . The two barometers 
( 1)1 have thought proper to mention this, as it is almofl the only circum- 
ftance wherein my method of obferving differed from. Mr. de lug’s, whofc 
thermometers (if I miftake not) were hung always in the Sun, and probably 
for this reafon, becaufe the column of the atmofphere between the two baro- 
meters, whofe mean heat is to be determined, is (if the Snn fliine) all expofed 
to the Sun. I have, however, always preferred hanging them in the fhade, 
and I give the following reafons : all fpurious and local heat from refle&ion is 
more eafily avoided ; no concentrated and falfe heat is acquired by the mounting, 
and thence communicated to the tube, even though the ball fhould be infulated; 
and, finally, becaufe I fufpeft the real temperature of the atmofphere in the 
Sun and in the fhade to be the fame, or at leaf! infenfibly different. This may 
be thought to be advancing too much; but, to be fatisfied of the pofition, I made 
no lefs than four-fcore obfervations with four different thermometers of very 
different mounting, hung alternately expofed to the Sun’s rays, and fereened 
from them by the fhade of a tree, in an open plain at fome diftance from the 
town of Geneva. The refult was, that my beft thermometer, with the ball 
infulated, differed only 2°’ in the different fituations; the others, more or lefs, 
as they were more or lefs conne&cd with the frames in which they hung. One 
of them, inclofed in a glafs tube, rofe 12° higher than the true temperature, 
which was 7 7 0 . It fhould feem then, that the variety in the mounting occa- 
fioned this difference; and this effedt of the materials, of which the inflru- 
ment is made, cannot be wholly avoided, as the glafs itfelf, which conflitutes 
the ball of the thermometer, will acquire and contain more or lefs, in proportion 
to its thicknefs and opacity. If a thermometer were perfeft, it would refleft 
all the rays that it receives. More might be added to corroborate this idea, 
but it would fwell this note to an unwarrantable length. 
were 
