INCANDESCENT MANTLES. 401 



alwa3\s been a weakness in the fabric, and to make the strangulation at 

 the top of the mantle l)y sewing it around with fibers of the same charac- 

 ter as those of which the mantle itself is made, this constricted annu- 

 lus then resting upon a supporting ring which is fixed as the ordinary 

 support in the center of the mantle. 



The asbestus thread has always weakened the top of the mantle 

 owing to the difference in the rate of contraction during burning off, 

 and with a mantle merely resting on a supporting ring in this way you 

 get what is really an antivibrating arrangement which materially 

 enhances the life of the mantle for outside work. 



The matter which 1 have brought before you has already occupied 

 so much time that 1 do not propose to go into the question of burners. 

 Many highly vaunted improvements have been introduced during the 

 last two years, but they have all been based upon the idea first intro- 

 duced ])y Bandsept of getting the proper admixture of the maximum 

 of air and gas to be burned at the bottom of the burner and completing 

 and perfecting the mixture close to the mouth of the burner where 

 the combustion is to take place. 



In many cases the increased luminosity given by such burners is 

 really due to the length of tube, the increase of which acts in the same 

 way as increase in gas pressure. These improved l)uruers have shown 

 themselves to be extremely variable, as while with carefull}' adjusted 

 samples and careful manipulation in the laboratory it has been possi- 

 ble to get as much as twenty-five candles per foot of gas with a prop- 

 erly prepared Welsbach mantle, yet in practice on a big scale the duty 

 given is quite as often seventeen candles or less, and the old Bandsept 

 burner, when properly made, is still as good as or better than any of 

 the new ones. 



