204 MUNROE— EXPLOSIONS OF PETROLEUM VAPOR. [April 22, 



the bed of Cork Run, and its laterals those of the tributaries to 

 Cork Run. The total length of the sewerage system thus con- 

 structed to the point where the collision occurred was 2,785 feet 

 and the different sections varied in diameter from 2 feet at the 

 head to 10 feet at the mouth where the sewer opened into the unfilled 

 portion of the ravine near the Ohio River. 



Immediately beside and parallel with the Ohio River was the 

 road bed of the P. & L. E. R. R., a culvert having been built over 

 Cork Run. Immediately beside this railroad embankment, but be- 

 tween it and the bluff was a turnpike road including its wooden 

 bridge across Cork Run, this bridge being 19 feet above the bed 

 of the run. Cork Run from the mouth of the sewer to the point 

 of its discharge into the Ohio River, a distance of about 116 feet, 

 was practically an open drain. The building of the tunpike across 

 the ravine made a large pocket or basin at this point. 



Naturally use was made of Cork Run sewer in draining Sher- 

 aden Yard, and, as a part of this system, catch basins for surface 

 water from the yard were built and connected to the sewer, one of 

 these catch basins being near the point at which the collision oc- 

 curred. 



Among other explosions, one occurred in the basin between the 

 mouth of the sewer and the turnpike which was sufficiently violent 

 to wreck many buildings in the vicinity and to lift from its track 

 a trolley car, which was just approaching the turnpike bridge, with 

 sufficient force to injure some of the passengers. 



Among other theories, it was alleged that this explosion was 

 due to naphtha which ran into the sewer through the catch basin 

 above referred to, where its vapors formed, with air, a combustible 

 mixture; that the vapors in the sewer were ignited by the fire 

 burning in the yard about the mouth of the catch basin ; and that 

 the flame was transmitted through the sewer and caused the ex- 

 plosion of the accumulated air-vapor mixture in the basin. 



In considering the probability of this theory being the correct 

 one, the dimensions of the sewer throughout its complete length 

 was ascertained and, from the volumes calculated from this data, 

 it was found that 3,451 gallons of naphtha would furnish sufficient 

 vapor to completely fill the sewer, while 200 gallons of the naphtha 



