^54 Ohfcrvhlhns on the PreiKing of IVatery 



tion, the fpoiijiy appearance which water acquires bv freezing 

 in vacuo, in coniequence of its greater expanfion, which every 

 time that the glals biirit was iinaccompanieci with the appear- 

 ance of air-hubbles, but inimciliatcly on the introdu<^tion of 

 air thev fpread tlicnifolves through the whole mafs. Hence 

 tliere was no doubt that this increafcd dilatation was entirely 

 the confcquciice of the aeriform wiUer^lhe expanfion of which 

 was not impeded by the preflure of the atmoiphere. We re- 

 peated this experiment for the la(t time on the fame day, 

 when th<j frolt was perceptibly lefs intenfe. A beer-glafs 

 capable of containing ciglu ounces was two-thirds filled with 

 di (tilled water as much freed from air as poflible : it was im- 

 inediatcly placed under the receiver of an air-pump, and the 

 air was exhaulied as fpcedily a fpoflible. At the end of fome 

 hours we faw a phtenomenon which appeared fo (trange and 

 fiuffular, that I was at lirlt incapable ot ia\ ing what it might 

 be. The freezing water had raifed ilfelf up in a ftraight 

 column nearly three inches above the glafsj being in thick- 

 ncfs in the middle above a line, and at the bottom two lines. 

 The water was equally frozen at the furface and throughout 

 the whole mafs : the cellular or Ipongy appearance exifted 

 onlv under the column, and extended to the middle of the 

 bale. No air-bubbles were any where to be feen. It is pro- 

 bable that the tlownefs of the freezing may have contributed 

 to produce this fingular phsenonienon. 



We conjcdnred, and with great probability, that this co- 

 Imnn was hollow, and had ferved for difchargmg the aeriform 

 water. As foon as the air was admitted under the receiver 

 we faw a movement in the pipe like that of air which pene- 

 trates into a wet glafs tube, and in the mean time a great 

 nianv air-bubbles appeared in the cavities at the bottom of 

 the column. Bv introducing a (liif bridle into the aperture, 

 and employ ine* a magnifyinsr clafs, we could eafily trace the 

 hollow part of the lube to the cavity where the air-bubbles 

 were. 



The air, therefore, on this occafion had no fiiare in the 

 expanfion of the ice ; and as our water before freezing never 

 contains fo much air, it appears that during the freezing it 

 could contribute little or nothing towards the expanfion. 

 " Expanfiiin is a pha?nomenun common during all cryftal- 

 lization, and is a confequence of the peculiar regular form 

 which the nioleciilce of bodies adhering mutually to each other 

 afhime, and by which means larger vacuities are formed. 

 Expanfion is increafed, and particularly during the cn'ftal- 

 lization of water, bv the aeriform ftale to which a yartof it 

 is brouffhl bv the difenff{»gt*d ealoric. 



^ ' ^ While 



J 



