[ 4g 5 ] 
tion was divided into four roundifh clumps, or 
clufters, which were feparated or diftinguifhed from 
each other by a furrow. Each clump was made 
up of a number of very fine delicate plates laid 
edgeways, in fomewhat of a regular manner; and 
between them a number of others, where part of 
the flat fides were to be feen, and amongif them an 
infinite number of fmali rhomboidal or roundifh 
cryftals ; the clumps appeared, in miniature, in the 
fun, fomewhat like to the lower part of the fpread 
tail of a peacock. The letters c, c, c , c , fhew 
the general form of the cryftallifation, d, d , d, the 
form and fihape of fome of the feparate cryftals.. 
I treated another parcel of plums of the fame 
kind in a different manner; I faturated part of the 
liquor when frefh, and let it ftand till a fermenta- 
tion had taken place before 1 evaporated it ; and I 
let the other half ftand till the fermentation was 
over before faturating it; but the fait obtained from 
both, appeared nearly in the fame form, though the 
number of clumps in this fecond cryftallifation was 
only three : and in a third experiment the appear- 
ances were exactly fimilar, only the fait did not 
divide into clumps. 
This fait taftes coolifh on the tongue, but does 
not affedt the thermometer, in the time of its folu- 
tion in water. 
In the prelent hard froft* fome of the fait of plums, 
which had been diflolved in three or four times its 
own weight of water, and let by in a clofet, cryftal- 
lifed anew. The cryftals were flat, thicker than a 
Ihilling, and moft of them had fix fides of unequal 
* N. B. The account of this laft part of the experiment was 
given in to the Royal Society in the beginning of January 1 768. 
lengths., 
