VOL. LXXIII.] PHILOSOPHICAL TRANSACTIONS. 415 



yet contain fluid quicksilver; but he thought it was thicker than ordinary, as it 

 did not run freely, but seemingly in pieces, not globules : however he put it 

 back again into the mixture, and set it by as of no further use ; but returning 

 after breakfast, he found it was firmly frozen, so as to give no appearance of 

 fluidity, though the included thermometer was only at 40°, which he considers 

 the exact freezing point of quicksilver ; and that the congelation was in fact 

 begun before, and effected by only a longer continuance in the same degree of 

 cold. 



Mr. H. made a 4th experiment Jan. 7> 1782. This was with the mercurial 

 thermometer a, and the apparatus f, as in the 1st and 2d experiments. The 

 day was clear, with little wind at w. by s. or w. s. w. which is generally the case 

 in this country in the coldest weather. The thermometers at 8 o'clock were as 

 follows, according to the rotation of the letters from a to g, 3() i, 36%, 35°, 

 25°, 25°, 34°4-, 34° below the cypher. The apparatus thermometer f, after 

 standing at 42° and 41°i for a considerable time, sunk at once to 77°, not 

 gradually, but suddenly as a weight falls. The great descent of the quicksilver 

 in the index thermometer a to 440° in the first freezing mixture he imputed to 

 the coldness of the weather, but was surprized to find it did not sink more than 

 10° lower in the 2d mixture. It is remarkable, that after pouring in the first 

 mixture on the 2d, the apparatus, which had risen a little before, sunk suddenly 

 into the bulb. At ll h 21 m Mr. H. took the apparatus out to examine it, and, 

 by shaking it in his hand, all of a sudden some of the quicksilver in the cylin- 

 der liquified ; the concussion perhaps dissolved its solidity, for it was not above a 

 minute out of the mixture. Wondering much at this unexpected phenomenon, 

 as the quicksilver in the thermometer did not rise, he put it into the mixture again 

 immediately ; but finding the inclosed thermometer showed no alteration, his 

 curiosity determined him to examine it again ; therefore, about 4 minutes after, 

 he took it out a 2d time, and found the surface of the quicksilver in the cylinder 

 was liquified about £ of the whole quantity ; the rest formed a solid ball, in- 

 cluding the bulb of the thermometer, which easily accounts for the quicksilver 

 in that instrument remaining stationary. 



A 5th experiment was made Feb. 22, 1781. The weather was clear and 

 serene, the wind about s. s. w., and the several thermometers stood as follows, 

 A 82°, b 66°, d 34°, e 34°, f 42°, g 42°, h 46°, at 7 in the morning, and at 

 8 o'clock they were a 78°, b 1 14°, d 2g°i, e 2Q%, f 29°i, g 40°, h 43° ; yet it 

 is remarkable, that quicksilver which was constantly exposed to the air in a 

 saucer was not frozen. Mr. H. imputes the small descent of the quicksilver in 

 the thermometers to the great degree of the cold in the atmosphere as in the 6th 

 experiment, for there the effect was similar. The most remarkable circumstance 

 in this day's operation was the sudden descent of the quicksilver in the apparatus 



