▼OL. LII.] PHILOSOPHICAL TRANSACTIONS. 547 



before he immersed his thermometer in it: he then, in like manner, put the 

 snow to the aquafortis, before he put his thermometer therein. In whichever of 

 these ways he proceeded, he found the event exactly the same; as the whole de- 

 pended on the aquafprtis dissolving the snow. When he had proceeded so far as 

 to find the mercury immoveable, he broke the bulb of the thermometer, which 

 had already been cracked in the experiment, but the parts were not separated. 

 He found the mercury solid, but not wholly so, as the middle part of the sphere 

 was not yet fixed. The external convex surface of the mercury was perfectly 

 smooth : but the internal concave one, after the small portion of mercury which 

 remained fluid, was poured out, appeared rough and uneven, as though com- 

 posed of small globules. He gave the mercury several strokes with the pestle of 

 a mortar which stood near him. It had solidity enough to bear extension with 

 these strokes ; its hardness was like that of lead, though somewhat softer; and 

 on striking it sounded like lead. When the mercury was extended by these 

 strokes, he cut it easily with a penknife. The mercury then becoming softer by 

 degrees, in about 12 minutes it recovered its former fluidity, the air being then 

 197. The colour of the congealed mercury scarcely difl^ered from that of the 

 fluid; it looked like the most polished silver, as well in its convex part as where 

 it was cut. 



The next day the cold had increased to 212 degrees, which was 7 degrees be- 

 yond what it had ever before been observed at Petersburg. The season so much 

 favouring, he thought it right to continue his pursuit, not only in further con- 

 firmation of what he had already observed, but to investigate new phenomena. 

 In two thermometers he observed the same facts in regard to the congealing 

 of mercury, as he did the preceding day. In the bulbs which he broke, the 

 whole of the mercury was not fixed, as a very small portion, much less than that 

 of the preceding day, continued fluid. He treated this mercury as he did the 

 former: he beat it with a pestle, he cut it, and every thing was thus far the same. 

 But he saw a very great difiference in respect to the descending of the mercury in 

 the thermometer, the like of which did not occur to him, neither in the former 

 nor any of the subsequent experiments. From the former ones it appeared that 

 the mercury in the first experiment had only descended to 47 O when it became 

 immoveable, though the glass bulb was not cracked. In the experiment of the 

 25th it descended to 530; and in two thermometers on the 26th, to 050. But 

 as well in the thermometer which he used on the 25th, as in two of the 2()th, 

 the bulbs were cracked in the experiment: they cohered however; nor was the 

 least part of the bulb separated, but the congealed mercury seemed to aatiere to 

 all parts of the bulb. In the following experiments, he invariably found that 

 the mercury sunk lower, if the whole of it was congealed, than if any part of 

 it remained fluid. It then generally descended to 680 and 700, but the bulbS 



4 A 2 



