THE LIME-LIGHT 55 



bag, and unconsciously kept balancing the decrease of oxygen 

 by turning off the hydrogen tap, to keep at its best the waning 

 light, when bang ! There was an explosion almost as loud 

 as a pistol-shot in the nearly empty oxygen bag, rolling off 

 the weights on to the floor, and tearing the boards apart at 

 the hinges. There being hardly any oxygen pressure to resist, 

 the hydrogen had gradually forced itself back into the last few 

 cubic inches of oxygen ; and as there was hardly any back- 

 pressure at the nipple to resist a * pass-back,' the result was 

 the explosion. Owing to the small quantity in the bag, no 

 other damage was done, the bag itself not being injured, and 

 being often used afterwards; it expanded enough without 

 rupture, only the hinges giving way from the suddenness of 

 the shock. 



Lesson. Never work a bag down quite empty with the 

 mixed jet. Observe how low the boards come down when 

 empty, and always leave off while there is an inch or two to 

 spare. Also, so soon as you have to be frequently turning off 

 the tap of one of the gases at the jei to restore a fast de- 

 creasing light, it is a proof that the pressure is diminishing 

 rapidly in the other bag ; if you go on, mixture may occur, 

 and it is time to stop. With decently competent management 

 in other respects, this is the grand rule of safety, but I have 

 never seen it stated in any lantern manual. 



80. Double Pressure-Boards. To resume practical details, 

 the last step to safety was the abolition of all difference in 

 weights by the use of double pressure-boards as shown in 

 fig. 33. By this arrangement one bag is placed over the 

 other, under the same weights, and it no longer matters what 

 these weights are, so long as they are sufficient to keep up 

 pressure. 



In making these boards, the hinged ends of the two 

 outside ones must be wider apart. The two bags are not 

 placed in contact, because full bags would slip ; but either a 

 middle board, or a piece of stiff sail-cloth is connected to the 



