1885.J en Accidental Exjjlosions by Non-explosive Liquids. 229 



have been cracked, the principal portion of its contents having leaked 

 out into the bottom of the ship. The cracked jar was handed iij) to 

 the lower deck with the siccative still leaking from it, and orders 

 were given to throw it overboard on account of the bad smell which 

 it emitted ; this was done within a very few minutes after the jar had 

 been removed, and the first explosion occurred almost directly after- 

 wards. Instructions had been given to clear up the leakage from the 

 jar after the hatch of the mast-hole had been left off a little time, and 

 it appeared that a naked candle had been given to the man who handed 

 the jar up out of the small store-hold described by that name. There 

 appears very little room for doubt that an exj^losive mixture of the 

 vapour and air had not only been found in the j)articular space where 

 the jar was kept, but tliat it had also extended through the air-spaces 

 at the bottom of the shijD towards and underneath the powder- 

 magazine, so that even the air in the latter may have been in an 

 explosive condition, as many hours had elapsed between the time 

 when the smell of the petroleum spirit-vapour was first noticed and 

 when the first explosion occurred. 



The special committee which had inquired into the jDossibility of 

 the occurrence of a violent gas explosion in the coal-bunkers of the 

 Doterel, was directed to institute experiments with a view of ascer- 

 taining whether the vapour evolved by this xerotine siccative would, 

 in the circumstances indicated by the official inquiry, have furnished 

 an explosive gas-mixture possessing sufficient power to have produced 

 the effects resulting from the first explosion on the Doterel, and 

 to have exploded the powder magazine. A preliminary experiment 

 showed that when a small quantity of the liquid was spilled at one 

 extremity of a wooden channel 7 feet long and 2*5 inches by 3 inches 

 in section, the vapour had diffused itself in the space of three minutes 

 throughout the channel to such an extent, that, on a light being applied 

 at one end, the flame travelled along very rapidly to the other end, 

 igniting a heap of gunpowder which had been placed there. Some of 

 the liquid was. also spilled upon the bottom of a very large sheet-iron 

 tank, and after this had remained closed for about twenty-four hours, 

 being exposed on all sides to the cool air of an autumn night, and there- 

 fore not under conditions nearly so favourable to evaporation as those 

 obtaining in the hold of a ship, the application of flame produced an 

 explosion of such violence as to tear open the tank. Experiments 

 were also made with the liquid in an old man-of-war, under conditions 

 somewhat similar to those which existed in the Doterel, and destructive 

 effects were obtained of a nature to warrant the conclusion that the 

 first explosion in the Doterel might have been due to the ignition of 

 an explosive mixture of the air in the confined space at the bottom of 

 the ship, with spirit vapour furnished by the liquid which had leaked 

 out of the jar. 



It is very instructive, as indicating the manner in which volatile 

 liquids of this class may, if their nature be unsuspected, be the causes 

 of grave disasters, to note that, while stringent regulations apply and 



