and on the Nature of Show. 251 



ceptacle for a verv larsie qnanlity of combined cakiric, which 

 fev the decoinpoiition of the atuiolphere returns attain, in a 

 great nifalore, with the water, as fenfible heat to the furface 

 of the earth : in the phcenon^cna of conibuftion, rcfpiration, 

 and other great operations in the ceconomy of nature, I can 

 i'ee not only a deconipofilion of vital air, but tliat alio of aiiri- 

 lorm water, which is h intimately coaneC:ted with air. But 

 as t dare not venture to proceed further in this path, fjar 

 which I yet feel myfelf too weak, I fliall reuirn to the prin- 

 cipal objeA of this paper. 



I have repeated tlie experiments wh/Kli I before coriimtmi- 

 cated to you, as already obl'evved, and with the (anie refult: 

 I entertained no doubt that the dilatalion of freezing water, 

 previouliv deprivetl of air, is to be afcribed, befides to cryf- 

 taliization, in particular to the aeriform water, which during 

 tlie frte^ing being carried oft" with the liberated caloric, de- 

 ranoes in its paOaji'e and feparates the cr\ Irals from each 

 other. I had entirely si'wtn up the idea of the exilience of 

 air in chemical uiiion with water, or with its bafes, which 

 could not be feparaicd by boiling, and vidvich bv freczinc* 

 might again be libtraied in an elaftic ftate, till the treatife 

 of mv friend Dr. Uilkens, written in March 1800, came into 

 my hands. 



The well-known merit of this philofopber, tlie modeftv 

 with which he communicates bis ideas to the public, mult. 

 me confidtT it as mv dutv to examine his ubjcftions widt a\\ 

 due relpect ; but at the fame time to prove to my pupils, if 

 poffible, the truth of my experiments. . / 



According to mv experiments, water can be perfeftlv freed 

 from air. In fuch water no air-bubbles ariie during the pro- 

 cefs of freezinsT, provided the air has no accds t^) it. Its 

 communication with the air would inllantly make them ap- 

 pear. 



For the fake of thofe who may be dtfirous to repeat my 

 experiments, f fliall delcribe an infallible method by whiiil 

 the fmalleit traces of air in water may be detected. None of 

 the meanp commonlv nfed for the purpofe aie futficient wheu 

 accuracy is rcjuired, as is the caie in the prcfeut inttance, 



I pour the water deliimd ior examination into an earthen 

 veflel, communicate a blue tint to it by a futficicnt quantity 

 of folution of litnui'S, wiiich I preferve ready, of a certain 

 llrength, for fuch purpoles: I ihtn fill with it a flafk in fuch 

 a manner, that the npcrlure of the flalk is kept under water 

 during ihc time of fiHin<r; taking care to (kd<e the fialk, that 

 the futallcd air-bi)bMo mav not remain in it. I then invert 

 tlie flaik in ihe water, aind iiitpod^icu into it, iu the comniou 



manner. 



