EXPERIMENTS WITH THE FLAME-MANOMETER. 401 



of a common lamp cylinder, separated by pastille from the wider 

 part, may be used for this purpose ; the tube should not be wider or 

 longer, otherwise it may begin to sound like a chemical harmonicon. 



The images of the flame are somewhat faint; the experiments 

 should therefore be performed during the evening in a dark room. 

 The flame is best fed by common gas ; if gas is not to be had, 

 hydrogen must be used. The flame of burning hydrogen is so faint, 

 that the images could scarcely be seen even in a dark room. It is 

 therefore necessary to pass the hydrogen, which is made by means of 

 the apparatus shown in fig. 156, (see p. 226) through the stopcock li 

 of that apparatus by means of an india-rubber tube into a small jar, 

 arranged as in the apparatus shown in fig. 154, the cotton wool in 

 which has been previously moistened with petroleum-spirit. The 

 vapour of this volatile liquid mixes with the hydrogen which passes 

 through the jar, and renders it capable of burning with a strongly 

 luminous flame. If needful the apparatus, fig. 154, may be used 

 for the whole arrangement, including the generation of hydrogen, 

 simply moistening the cotton wool in the small jar by the petroleum- 

 spirit. But, in that case, it often happens either that the flame is 

 too small, as when too little sulphuric acid is poured into the genera- 

 ting bottle, or that it is too large, when too much of the acid has 

 been used, in which latter case the liquid is apt to rise in the funnel 

 tube and to run over. 



The handle of the whirling table is turned quite slowly, so that 

 the box may rotate about once in a second. Instead of working the 

 handle, the box may be turned by the finger, placing the tip upon 

 one of the upper corners. 



The india-rubber tube which is attached to the tube a should have 

 a width, if possible, of 8 mm , at any rate of not less than 6 mm ; its 

 length may be from O3 to l m . A small funnel is inserted into the 

 free end, and notes are sung or whistled into it ; or the end of the 

 tube may be simply placed lightly between the teeth, without 

 applying any pressure, and the notes may be sung, the mouth being 

 half open, or the mouth may be closed and the note hummed into 

 the tube. 



The rotating mirror-box may also be used for investigating the 

 flame of the chemical harmonicon, which gives also a series of 

 images side by side. 



The appearance of the images of the flame, shown in 

 fig. 236, is that produced by the action upon the flame 

 of a simple tone, that is, one not accompanied by per- 



D D 



