CHEMICAL HISTORY OP A CANDLE 



101 



but you can, no doubt, by this time generalize enough to 

 be able to compare one thing with another : what I am about 

 to do is to change the ascending current that takes the flame 

 upward into a descending current. This I can easily do by the 

 little apparatus you see before me. The flame, as I have said, 

 is not a candle flame, but it is produced by alcohol, so that it 

 shall not smoke too much. I will also color the flame with 

 another substance, ( 6 ) so that you may trace its course; for, 

 with the spirit alone, you could hardly see well enough to 

 have the opportunity of tracing its 

 direction. By lighting this spirit 

 of wine we have then a flame pro- 

 duced, and you observe that when 

 held in the air it naturally goes up- 

 ward. You understand now, easily 

 enough, why flames go up under 

 ordinary circumstances: it is be- 

 cause of the draught of air by 

 which the combustion is formed. 

 But now, by blowing the flame 

 down, you see I am enabled to 

 make it go downward into this 

 little chimney, the direction of the 

 current being changed. Before we have concluded this course 

 of lectures we shall show you a lamp in which the flame goes 

 up and the smoke goes down, or the flame goes down and 

 the smoke goes up. You see, then, that we have the power 

 in this way of varying the flame in different directions. 



There are now some other points that I must bring before 

 you. Many of the flames you see here vary very much in 

 their shape by the currents of air blowing around them in 

 different directions ; but we can, if we like, make flames so 

 that they will look like fixtures, and we can photograph them 

 1 indeed, we have to photograph them so that they become 

 fixed to us, if we wish to find out every thing concerning 

 them. That, however, is not the only thing I wish to men- 

 tion. If I take a flame sufficiently large, it does not keep that 

 homogeneous, that uniform condition of shape, but it breaks 



The alcohol had chloride of copper dissolved in it: this produces a bean 

 gtful green flame. 



FIG. 59 



