358 



SCIENCE. 



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



of one gas back into the reservoir of another 

 gas, but this is entirely prevented by proper 

 regulation of the pressure and size of the 

 orifice. Another arises from confusing the 

 cylinders when filling them, and to prevent 

 this the cylinders have been painted differ- 

 ent colors. Yet, as shown by the fatal ac- 

 cident described by W. N. Hartley,* this 

 has not prevented the deliberate interchange 

 of the cylinders under the pressing demands 

 of trade, and the usual causalty has followed. 

 Therefore, he proposes that the fittings for 

 the two classes of cylinders be made so en- 

 tirely different that it will be practically 

 impossible to charge the cylinder with the 

 wrong gas, and in view of the probable in- 

 creased use of gas in this form, as indicated 

 by Mr. Thomas Fletcher ,t the change should 

 be made. Yet I doubt if it will be, except 

 under compulsion of law, for I have learned 

 in my efforts to introduce safety explosives 

 in this country that the great majority will 

 not secure the assurance of safety if this 

 entails a little inconvenience and the taking 

 of a little more pains. 



A more common source of accident has 

 come from impurities introduced in the 

 making of the oxygen, as, at Nabant, Mass., 

 where pulverized stibnite was mistaken for 

 pyrolusite, and mixed with the potassium 

 chlorate. Limousin describes an accident 

 at Cannes in 1880, which attracted unusual 

 attention from the factitious circumstance 

 that the gas was being prepared for the 

 Empress of E,ussia,;|; and found the cause 

 in the evolution of hydrocarbons from the 

 rubber connecting tube by particles of 

 heated potassium perchlorate carried into 

 it through the turbulence of the reaction. 

 While Professor C. A. Young gives an ac- 

 count § of the explosion at Princeton while 



* Chem. News, 59, 75 : 1889. 



t 'Ou a New Commercial Application of Oxj'gen.' 

 J. See. Chem. Ind. 7, ISS : 1888. 

 i U. S. Nav. Inst. 14, 167 ; 1888. 

 JSci. Am., p. 369, June 11, 1887. 



filling a steel cylinder with oxygen by 

 means of a water-jacketed, steam force 

 pump, and finds the cause in oil used for 

 lubricating the pump being sprayed into the 

 gas cylinder so as to form an explosive 

 mixture with the oxygen. He recommends 

 the use of soap suds as a lubricant in place 

 of oil. Frankland* describes a similar in- 

 stance and gives a similar explanation. 



Recently my attention has been called to 

 several accidental explosions of oxyhydro- 

 gen mixtures formed in the operation of 

 storage batteries, the detonating gas being 

 fired by the spark formed on breaking con- 

 nections at the battery. 



But of all circumstances under which 

 explosions occur the most awful are those 

 which so frequently happen in mines, for if 

 the miner escapes instant death it too often 

 is but to die from suffocation, or, worse yet, 

 to be entombed and perish from starvation 

 preceded perhaps by insanity. 



It has long been known that fire damp 

 found its way into coal mines, and in 1674 

 Mr. Jessop communicated to the Royal So- 

 ciety a description of the accident met with 

 by Mr. Michel, who penetrated into the gal- 

 lery of a coal pit, in Yorkshire, with a 

 naked torch and was severely burned. It 

 is interesting to notej that, when rescued, 

 he declared he had heard no noise, though 

 the workmen in the vicinity had been ter- 

 rified by a tremendous report accompanied 

 by a vibration of the earth. As is to be 

 expected, from what we now know of nat- 

 ural gas, inflammable gases are not confined 

 to coal mines, but, as shown by B. H. 

 Brough,! they are met with in metalliferous 

 mines and other excavations also. 



The appalling nature of these catastrophes 

 led to efforts being made to at least reduce 

 their frequency, if not to prevent them alto- 

 gether, an extended account of these being 



* Am. Gas Light J. 5, C'SO ; 1864. 



t Treatise on Coal Gas, Wm. Richards, p. 4, 1877. 



t School of Mines Quarterly 12, 13-SS ; 1890. 



