628 Mr. U. Behn : Some new 



escape from the holes) and is then lit at the openings. The 

 flow is adjusted so that there is a small luminous tip to the 

 flames, the tube being meanwhile supported in a horizontal 

 position. 



If the tube is now slightly tilted, so that one end, say A, 

 is higher than the other, it will be found that the flame at 

 the higher end of the tube becomes comparatively large and 

 that at the lower end comparatively small. The same effect 

 is obtained, whichever end of the tube is raised, so it cannot 

 be caused by inequality in the size of the holes, or by 

 friction in the tube. (At A the gas-pressures inside and 

 outside the tube are sensibly equal, as is shown by the small 

 size of the flame. Above A the pressure decreases more 

 quickly in the dense air than in the light coal-gas. Hence 

 at B the pressure inside the tube exceeds the pressure 

 outside, and so the flame is large.) If one end lies only 

 three millimetres higher than the other, a difference in the 

 size of the flames will be produced, which, if the air is 

 undisturbed, may be plainly seen from all over a large 

 lecture-hall (15-20 m.) . In this way a difference in pressure 

 of about (*0012- , 0006) '03 gram weight per sq. cm., or 

 •000,000,2 atmosphere * is made visible. 



A new field of application for the flame-tube lies in its use 

 for indicating small motions, or rather small accelerations. 

 A slight movement of the tube causes changes of pressure 

 in gas, due to the latter's inertia, and these changes of 

 pressure in the tube are made visible by the difference in the 

 sizes of the flames. For example, the beats of a seconds 

 pendulum, though they be very small, may be made visible 

 to the naked eye. For this purpose the flame-tube is 

 suspended like a pendulum. The tube for these experiments 

 need only be about *5 m. long f and one flame-hole at either 

 end is sufficient. On account of the experiments described 

 further on, the tube should be as light as possible. For the 

 experiment in question, one ought to increase its weight by 

 putting a f| -shaped piece of lead sheeting on it (fig. 2). 

 Oscillations of one cm. amplitude will suffice to make 

 the jumping of the flames visible from all over a lecture 

 theatre. 



* Approximately ; the difference of temperature of coal-gas and 

 surrounding air being not very large. 



t In experiments, for which, on the contrary, a very long flame-tube 

 is required, it is sometimes advisable to build the tube up by a number 

 of pieces put loosely together. In this way one may avoid the bending 

 of the tube, which of course is stronger heated by the flames from the 

 upper side. 



