I9I0.] MUXROE— EXPLOSIONS OF PETROLEUM VAPOR. 



205 



would fill the sewer with a slightly combustible but non-explosive 

 mixture. As the tank car, which was perforated beside the catch 

 basin, contained 7,253 gallons of naphtha, there was a sufficient 

 volume to more than satisfy the requirements for filling the sewer 

 with naphtha vapors as set forth above. 



That explosions occurred in or about portions of the sewer and 

 its laterals seemed undoubted, yet recalling the safety lamp, the 

 difificulty experienced in transmitting flames through tubes, and the 

 varying capacities of columns of explosives of dififerent diameters 

 in propagating detonation, a doubt arose in my mind as to the flame 

 traversing this sewer. Not finding any information on this matter 

 in literature, experiments were made as follows : 



Results of Experiments. 



It having been observed that if vessels were filled with the 

 combustible mixture of naphtha vapor and air and the mixture was 

 Ignited at the mouth of the vessel, the flame retreated to various 

 depths depending on the relation of the diameter of the vessel to 

 its length, experiments were made with tubes. In conducting the 

 experiments in each case the tube or pipe, which was open at both 

 ends, was inclined at an angle and a considerable quantity of liquid 

 naphtha was poured slowly into it at the upper end. So soon as 

 the last of the liquid had been poured into the tube a flame was 



