422 WRITINGS OP JOSEPH HENRY. [1856 



the cylindrical part was concealed behind the wainscoting, 

 and consequently the little image made its appearance only 

 in fine weather. 



A water-barometer was constructed by Professor Daniell, 

 and placed in the hall of the Royal Society, of which a full 

 account has been published in the Transactions of that insti- 

 tution. A minute account is given of the method of blowing 

 the tube, and the details of permanently fastening it in the 

 box which was to form the case. The tube was left open at 

 both ends; to the upper one a stop-cock was attached, and 

 the lower one was inserted in a small steam boiler, which 

 served the purpose of boiling the water to expel all the air, 

 of elevating it to the proper height by means of the elastic 

 force of the steam, and also as a permanent cistern to the 

 barometer. After the water was forced to the top and issued 

 from the stop-cock in a jet, the latter was closed ; the stop- 

 cock in the boiler was opened, steam suffered to escape, and 

 the water to settle in the tube until balanced by the pressure 

 of air. The upper part of the glass under the stop-cock, 

 (which had previously been drawn out into a fine tube,) was 

 gradually heated by a blow-pipe, and as soon as it was suffi- 

 ciently softened the pressure of the air effectually closed it. 

 The part above the stop-cock was then removed with a file. 

 This barometer was completed, after adjusting the scale, by 

 pouring a quantity of castor-oil on the surface of the water 

 to prevent contact with the air. 



After a series of observations however it was found in the 

 course of about three months that the column of water was 

 gradually descending, and it was finally resolved to open the 

 boiler and to examine the instrument. The oil upon the sur- 

 face was found to have undergone a change, though the water 

 below was perfectly bright and transparent. A portion of 

 the water was taken out and placed under the receiver of 

 an air-pump, and bubbles of air in abundance were extri- 

 cated; the air was absorbed by the water, diffused through 

 the whole mass to the top, where it was given off to the 

 vacuum, and thus caused the gradual descent of the column. 

 It was found however that it was not atmospheric air in the 



