184 



Dr. W. Marcet. On the 



constructed, the bell-jars being jacketed and the inside space of nearly 

 half an inch in thickness was filled with water. Later on I sub- 

 stituted for water a solution of sodium sulphate saturated at a 

 temperature exceeding that of my laboratory, so that the annular 

 space in the bell-jars became filled with a crystallised solution of the 

 salt. I thought by this plan to ensure a perfect stability of tempera- 

 ture in the air of the bell-jars during the analysis, any increase of 

 temperature outside being transformed into motion through its 

 action on the crystallised mass. The result was nearly as I had 

 anticipated; still, after trying the plan for a considerable time, I 

 finally gave it up and returned to the water-jacket which I now find 

 quite satisfactory. There is a scale to each of the bell-jars, divided 

 into four equal parts, and a pointer fixed to the rim of the tanks ; 

 by bringing the same line on the scale opposite the pointer 

 before and after the analysis, the operator will be certain that the 

 bell- jars hold exactly the same volume of air in both cases, admitting 

 of course that the atmospheric pressure has ' been perfectly re- 

 established. 



The jacketed air-holders are raised and lowered by means of an iron 

 weight worked over pulleys by a rack and drum movement, and 

 handle ; when the two bell-jars are at the same height, or level, they 

 may be entirely released of the weights without their position being 

 disturbed ; but if one of the receivers, say the left-hand one, is to be 

 depressed from that position and emptied of the air it contains, the 

 left hand weight is brought to bear upon it, and by a slow movement 

 of the handle that receiver is immersed entirely in the bath while the 

 other one is raised as high as it will go. When thus raised it draws 

 in common air ; this air is filtered through marbles or fragments of 

 pumice-stone, moistened with a solution of potassium hydrate, and 

 then through calcium chloride, so that no carbonic acid nor moisture 

 be admitted into the bell-jar. It is, of course, necessary to keep the 

 taps communicating with the external air open in both tanks when 

 the bell-jar is being emptied, or the oil will be drawn into one of the 

 stand-pipes. Should such an accident occur, which never happens 

 with a little practice, the oil will have to be let out from a small 

 aperture closed with a screw tap in the lower bend of the U~pip e ' 

 while the thermometer has to be removed from the bell-jar, letting in 

 air through the neck in which it fits. It is better to return an equal 

 volume of glycerine to the tank rather than oil, from the difficulty the 

 latter would have in finding its level in the tank. Should oil be added 

 the bell-jar will have to be raised so far as it can go, and time allowed 

 for the oil to assume its level below the rim of the receiver. 



The cord which connects the bell-jars is liable to expand or contract, 

 but this is remedied by turning a screw, placed so as to shorten or 

 lengthen the cord, while the eye is kept on the pointers of both tanks 



