March 10, 1899.] 



SCIENCE. 



351 



ber 21, 18SS, on the petroleum-laden ketch 

 'United ' at Bristol, England, through which 

 the docks were blown up, three men killed 

 and several injured, the glass in the win- 

 dows shattered for a radius of upwards of 

 300 feet, and extensive damage done by 

 fire. 



The accident was made the subject of a 

 special report by Col. V. D. Majendie* 

 which contains the results of his investiga- 

 tion and the experiments by Dr. Dupre and 

 Mr. Boverton Redwood, from which it ap- 

 pears that the material on the ' United ' 

 was ' deodorized naptha ' in forty-two gal- 

 lon barrels ; that the average annual leak- 

 age oa petroleum oil in barrels amounted, 

 in 1874, to 8 per cent, and on petroleum 

 spirit to double this quantity, and that, 

 though there has since been a great im- 

 provement in the treatment of the barrels, 

 it is still very large ; that one volume of the 

 liquid gives 141 volumes of vapor at ordi- 

 nary temperature having a specific gravity 

 of 3.5 to 3.8 ; that one volume of the liquid 

 will render 16,000 volumes of air inflam- 

 mable, 6,000 most violent explosive, 5,000 

 strongly explosive, and 3,000 scarcely 

 explosive but combustible. The naptha 

 vapor alone or when mixed with air in 

 the best proportions was not ignited by a 

 shower of sparks from flint and steel ; by a 

 stream of sparks from fireworks of various 

 kinds burning without flame ; by incandes- 

 cent match ends, or by incandescent plati- 

 num heated by electricity to a red heat. 

 Even red-hot coals held over and some- 

 times falling upon a small quantity of the 

 spirit spilled on a wooden floor failed usu- 

 ally to ignite it, and the cause in those 

 cases in which ignition did take place in these 

 red-hot coal experiments was uncertain, as 

 there was a fire burning in a near by room. 

 Ignition was, however, certainly effected by 

 the application of a flame or by contact with 

 a platinum wire approaching incandescence. 



* Eyre & Spottiswoode, London, 1889, 30 pp. 



The ' fireworks ' test makes a striking 

 lecture experiment, especially the one de- 

 vised by Mr. Redwood with ' vesuvians,' or 

 incandescent cigar lighters. For this pur- 

 pose he attaches two, of the glowing variety, 

 to a wire so that the tip of one will be in 

 contact with the base of the head of the 

 other. The latter is lighted, and when it 

 ceases to flame, and only glows, the mass is 

 thrust into the explosive mixture, where it 

 remains with the combustion, progressing 

 from tip to base and base to tip without 

 other effect until, when flame bursts from 

 the tip of the second vesuvian, the vapor- 

 ous mixture surrounding it is ignited and 

 an explosion ensues. 



Col. Majendie has properly called atten- 

 tion in this report to the fundamental dis- 

 tinctions between the danger arising in the 

 transportation of a cargo of dj'namite and 

 one of petroleum spirits, since in the former 

 case an explosion does not take place until 

 fire is brought to the dynamite, while in the 

 latter case the dangerous vapors will travel 

 to a fire at a considerable distance and even 

 through intervening bulkheads. 



For this reason mixed cargoes of which 

 volatile inflammable liquids and explosives 

 constitute a part are particularly dangerous, 

 as was long since shown in the explosion of 

 the canal boat ' Tilbury,' in Regent Park, 

 in 1874, having on board five tons of gun- 

 powder and four barrels of benzoline and 

 also having a small fire burning in the after 

 cabin some 35 to 40 feet from the forehold 

 in which the petroleum was stored. Not- 

 withstanding that the cargo was covered 

 with tarpaulins and that there was an in- 

 tervening bulkhead, the vapors I'eached the 

 fire and a most devastating explosion fol- 

 lowed. The cargo was thus made up in 

 spite of a similar disastrous experience 

 from similar causes on the ' Lottie Sleigh,' 

 at Liverpool, in 1864,* and neither of them 

 have proved a sufficient warning to alto- 



* Abel, loc. cit. 



