Experiments and Obfervations. 595 
I attempted to make thermometers with tubes of dif- 
ferently coloured glafs, but when a ball was formed with 
any of thofe tubes, the fubftance of the glafs in the ball, 
being much thinner than in the tube, differed very little 
from clear colourlefs glafs. 
To include the thermometers in clofe boxes, in which 
the rays entered through coloured glaffes, was alfo found 
ineffectual ; not only becaufe the colours fo tranfmitted 
were far from being homogeneous, but efpecially be- 
caufe fome of thofe glaffes are much more opaque than 
others, even of the fame colour. 
The leaft ambiguous method, therefore, was that of 
painting the balls of the thermometers with water-co- 
lours, taking care to lay them as equally fmooth as poffi- 
ble. In this manner I repeated feveral experiments, 
tiling fometimes a dozen of thermometers at once, 
whofe balls were painted with various colours, and were 
expofed to the Sun; and from a vaft number of experi- 
ments, and after fome weeks obfervation, it could be only 
deduced, that if the colours, with which the balls of the 
thermometer were painted, were pretty hke the prifmatic 
colours, thofe thermometers fhewed a greater degree of 
heat, whofe colours were nearer to the violet in the order 
of the prifmatic colours, and contrary wife; but they 
were all, even that painted with white lead, in fome 
4 1 intermediate 
