4\6 PHILOSOPHICAL TRANSACTIONS. [ANNO 1/83. 



thermometer, and the length of time it continued at 79° before the quicksilver 

 in the cylinder became solid. 



Experiment 6th was made Jan. II, 1782. After a cold night, the quick- 

 silver in the thermometer was at 44° below O at 7 in the morning : thinking this 

 great degree of cold was the most favourable opportunity of observing how low 

 it was possible to make the quicksilver descend in the tube of the thermometers, 

 Mr. H. resolved to embrace it, and at the same time to observe the concurrent 

 degrees with a spirit thermometer. The thermometers are marked from a to h, 

 and the observations are regularly in that order. At the beginning they stood as 

 follows, below 0, in the open air. 



A BCDEFGH 



At 7 h 45 444 45 41 28 29 40 40 46 

 7 50 40" 64 124 30 32 42 41 46" 



It is observable, that neither the quicksilver which was in the cylinders affixed 

 to f and g, nor the other quicksilver kept in the same place, some in a saucer, 

 some in a gallipot, and some in a phial, showed the least appearance of congela- 

 tion. Being engaged in preparing for the ensuing experiment, Mr. H. did not 

 remark either the great descent or ascent of the quicksilver in c, which must 

 have been very sudden, as his remarks are only 5 minutes asunder. The small 

 descent of the quicksilver in c, and the little effect produced by moving it into a 

 second mixture, made him at first apprehend the instrument was damaged ; he 

 did not however take it out, but took another thermometer a, and put it also in 

 the mixture; but he found it was stationary at a higher degree than c : he there- 

 fore exchanged a for the mercurial thermometer b, which to his great surprize 

 was stationary at S6°, nor could it be got lower till the cold of the mixture di- 

 minishing, it fell at once to 434°, and a few minutes afterwards c fell to 360°. 

 Imagining that a new mixture would now bring it very low, he made another, 

 but in the mean time the instruments had risen greatly, and after standing 

 in the fresh mixture, c sunk to 374°, and b to 438°. These mixtures were 

 double in quantity to those used in the former experiments ; instead of glass 

 tumblers, they were made in pint basons. 



Mr. H. observed also, that the mixtures seemed to grow thin sooner than 

 common ; for he always made them of the consistence of pap. He added snow 

 at times, to thicken it, but found it had very little effect, but rather decreased 

 the cold. While the instruments were stationary in the foregoing experiment, 

 he put the apparatus f and g severally into the mixture with the others ; the 

 consequence was, that in 1 minutes the quicksilver in the cylinder was frozen 

 solid. At g h 48 m put in apparatus f, when it stood in the air at 40° or 4 1° below 

 O ; and at 9 h 50 m took it out frozen solid, and the inclosed thermometer point- 



