in Comlustion, and in ike Condensation of Vapours. 2S9 



onder the aperture of the worm, and to cover the apparatus 

 on al! sides by screens, to prevent the flame from being 

 deranged bv the wind. 



There is in these experiments a source of errors too evi- 

 dent to escape ilie most superficiJil observation, and to 

 which it is indispensable to pay attention. While the ca- 

 lorimeter is heated bv the l-.eat developed in the combus- 

 tion of the inflammable substance which is burnt at the 

 aperture of its worm, it is continually cooled by the cir- 

 cumambient air. It would doubtless be possible, by calcu- 

 lations founded on a knowledge of the law of the cooling 

 of ihe receiver, which might be discovered by particular 

 experiments, to determine the measure of the effect produced 

 by the cooling in question, and even with a certain degree 

 of precision: but it would be impossible to appreciate by 

 this method, or ijy any other method known, the effects of 

 another cause of error,' less ostensible perhaps, but certainly 

 more powerful, than that of the cooling of the external sur- 

 face of the receiver. 



The aZiUe which is mixed with the oxygen of the at- 

 mosplieric air is necessarily carried into the worm, with 

 the products properly belonging to combustion; and without 

 a precaution which occurred to me to prevent the effects of - 

 this cause of error, by compensating for them, all my ex- 

 periments would have been useless. 



Fortunately the method which I employed to prevent the 

 effects of this cause of error, was sufficient to prevent at the 

 same time those which might have resulted from the cool- 

 ing of the external surface of the receiver. 



\s, the receiver is not cooled by the atmospheric air 

 which touches its external surface, or by the azote and the 

 other gases which pass through the worm with the pro- 

 ducts of combustion, unless The worm be warmer than 

 the surrounding air; and as on the contrary it is heated by 

 these same elastic fluids, always when its temperature is 

 lower than theirs ; bv taking care that the temperature of 

 the water in the receiver is alwavs at the commencement 

 of an experiment a certain number of degrees by the 

 thermometer (five for example) below the temperature 

 of the air, and by fiuislung the txperimcnt at the ins'ant 

 when the water in the receiver shall have acquirtii a hi^'her 

 temperature than that of the air of the snine number of de- 

 grees, the receiver will be heated by the air durmg halt the 

 lime occupied by the experunent, and cooled during the 

 other half: the calorific and frigoriiic effects of the air on 

 the apparatus will be counterbalanced so as not to produce 

 any sensible effect on the results of the experiment, and 

 cousocjucntly -u as to require no correction. 

 Vol. 41. No. ISO. y//jr27 1813. T When 



