352 



SCIENCE. 



[N. S. Vol. IX. No. 219. 



gether prevent subsequent reckless disre- 

 gard of all dictates of common prudence. 



Yet, because of these experiences, at- 

 tempts have been made in some instances, 

 where small lots of spirit were taken by 

 vessels, to avert disaster by carrying them 

 as deck loads, but the experience on the 

 ' Solway,' which carried 24 barrels of this 

 article on the main deck before the poop, 

 shows that this does not ensure security, 

 for, meeting with heavy weather, the casks 

 broke adrift, their vapors reached the galley 

 or cabin fires, and the vessel, with 19 per- 

 sons, was lost. 



Even where great precautions are taken 

 to prevent accidents they not infrequently 

 occur from inflammable substances being 

 met with in unexpected places, or being in- 

 troduced surreptitiously in admixture with 

 harmless bodies. Nowhere, perhaps, is more 

 care taken in this respect than on passenger 

 steamships and in the naval service, yet 18 

 years ago a series of accidents occurred on 

 board English ships, the cause of which 

 was for a time veiled in mystery and which, 

 in the then-existing state of feeling conse- 

 quent on the dynamite outrages, aroused 

 the gravest apprehensions. 



In June, 1880, a violent explosion took 

 place, without any warning or apparent 

 cause, in the forepeak of the Pacific Steam 

 Navigation Co.'s steamer 'Coquimbo,' 

 shortly after her arrival in Valparaiso. 

 Several plates were blown out of the bow, 

 and other structural damage was inflicted 

 while the ship's carpenter, who was the 

 only person apparently who would have 

 thrown any light on the cause of the acci- 

 dent, was killed. 



This explosion was followed on April 

 26, 1881, by a much more serious one on 

 the man-of-war ' Doterel ' (while at anchor 

 off Sandy Point, in the Straits of Magellan), 

 through which eight ofiicers and 1.35 men 

 lost their lives and the vessel was destroyed. 

 In May, of the same year, an explosion 



of trifling character happened on H. M. S. 

 ' Cockatrice,' in Sheerness Dockyard ; while 

 in November one, which was sufficiently 

 severe to kill two men, dangerously wound 

 two more (one fatally) and injure six oth- 

 ers, besides doing much damage to the 

 ship, occurred on H. M. S. ' Triumph,' then 

 at Coquimbo. 



The first suggestion as to the real cause 

 of these accidents was obtained in the in- 

 vestigation of that on the' Cockatrice,' when 

 it was developed that, just previous to the 

 explosion, a man went into the store room 

 with a naked light which he held close to a 

 small can, that was uncorked at the time, 

 and which contained a preparation recently 

 introduced into the naval service (as a 

 ' drier ' for use with paint) under the 

 name of Xerotine Siccative, and that this 

 largely consisted of a most volatile petro- 

 leum product. As it had been issued with- 

 out knowledge of this fact, instructions 

 were at once sent out by the Admirality 

 directing that it should be stored and 

 treated with the same precautions as tur- 

 pentine and other highly inflammable 

 liquids or preparations ; and these instruc- 

 tions had but recently reached the ' Tri- 

 umph ' when the accident narrated hap- 

 pened to her. Inquiry here developed the 

 fact that the explosion originated in the 

 paint room through bringing a lantern to a 

 compartment in which a leaky can of Sic- 

 cative had been stored, and following up this 

 clue the explosions on the ' Coquimbo ' and 

 ' Doterel ' were fully and definitely proved 

 to have been due to the presence on board 

 of this same substance ; while experiments 

 with the material showed that it was ca- 

 pable of producing all the destructive effects 

 observed, except, perhaps, in the case of the 

 ' Doterel,' where, from the two reports noted 

 and the other resemblances to the Regent 

 Park explosion, there was but little doubt 

 that the powder magazine was also exploded. 



Such accidents were not, however, con- 



