ipril 2, 1885] 



NA TURE 



5i9 



ignited, causing an outburst of flame which may alarm a nervous 

 erson and cause the dropping or overturning of the lamp. The 

 iccident which occurred in some apartments in Hampton Court 

 Palace, in December, 1S82, and gave rise to a somewhat alarm- 

 ing fire, appeared almost beyond doubt to have originated from 

 the employment by a domestic servant of a contrivance in which 

 petroleum spirit was used for heating water; but, as petroleum- 

 ■ used in the particular residence where the fire 

 actually occurred, public correspondence ensued regarding the 

 uncling the use of such lamps, although all which 

 were known to have been on the premises were forthcoming 

 after the fire and found to be intact. There was, at any rate, 



ace whatever adduced in support of an assumption 

 that the casualty was due to the explosion of a lamp, and other 

 instances might be quoted in which the breaking out of a fire, 

 or the destruction of or injury to life, which had evidently been 



y upsetting or allowing to fall a petroleum lamp, has 

 been erroneously ascribed to an explosion. 



There are, however, numerous casualties which have been un- 

 questionably caused by the occurrence of explosions in lamps, 

 and which have in many cases been followed by the ignition of 

 the oil, and the consequent loss of life or serious injury to those 

 in the immediate vicinity of the accident. Careful inquiries 

 have of late been instituted into casualties of this kind, and in 

 many instances the explosions have been distinctly traceable to 

 some immediate cause. In the great majority of cases they 

 occur some considerable time after the lamp was first kindled, 

 and when the supply of oil remaining in the reservoir has been 

 but small. Occasional examples of the reverse are, however, 

 met with. Thus, last spring, a man and his young son were 

 sitting at a table reading, his wife being also close at hand, when 

 a paraffin lamp, which had just been lighted, exploded, and the 

 room was at once set on fire by the burning oil which escaped. 

 The husband and wife fled from the room, both being slightly 

 injured, but the child was unable to escape from the flame, and 

 was burned to death. The oil used in the lamp was of a well- 

 known brand, having a flashing point ranging from 73' to S6 3 F., 

 and assuming that the recently lighted lamp had been filled with 



was untouched at the time of the explosion, no satis- 

 factory explanation can be given of the accident, unless, perhaps, 

 the reservoir had been so completely filled with oil, that the 



m of the liquid, on its becoming slightly warm, exerted 

 force to determine the fracture of the glass at soaie 

 part where a flaw or crack existed. 



A lamp accident which occurred last July at Barnsbury, 

 causing the death of a woman and her husband, appears, on the 

 other hand, distinctly traceable to the production of an explo- 

 sion in the reservoir of the lamp. The latter was stated to have 

 been alight but a short time, when, the husband being already 

 in bed, the wife, in her night-dress, attempted to blow out the 



the lamp ; the man heard a report, and, looking towards 

 the lamp, saw his wife in flames. He proceeded at once to her 

 sverely burned in extinguishing the flames in 

 which she was enveloped. The woman died in a few hours, and 

 the man succumbed thr;e days later to the injuries received. 

 There being no witness to the accident, there is no evidence 

 against the supposition that, on the occurrence of a slight explo- 

 it: in the reservoir in the lamp, the woman, having hold of 

 it when attempting to blow it out, may have upset it, or 

 tilled it so as to cause the oil to flow out and become inflamed. 

 The lamp may have become fractured by the explosion ; but 

 whenever such a result has been produced, the lamp had always 

 been burning some time, so that there was considerable air-space 

 which could be filled by an explosive atmosphere, whereas, in 

 this case, the evidence appears positive as to the lamp having 

 been full of oil when lighted. 



In another fatal case of a lamp explosion in the same month, 



I iid, the accident was also caused by the attempt on the 

 part of a woman to blow out the lamp before going to bed. In 

 this case the lamp had been burning for three hours ; the husband 

 of the sufferer was in bed asleep in the room at the time, and, 

 m being unable to give any account of the occurrence, 

 the only information elucidating it was furnished by the 

 daughter, to the effect that the lamp had been burning for three 

 hours, and that it was the habit of her m ither to extinguish the 

 lamp by first lowering the wick and then blowing down the 

 chimney. 



Another fatal accident, caused by the explosion of a lamp, 

 took place at Camberwell last January, and was brought about, 

 as in the two preceding cases, by attempts to extinguish the 



lamp by blowing down the chimney. The husband and two 

 sons of the sufferer were witnesses of this accident ; the lamp 

 had been burning for six or seven hours, when the woman took 

 it in her hand, and having partially turned it down, proceeded 

 to blow down the chimney ; an explosion at once occurred, the 

 glass reservoir was broken, and the inflamed oil flowed upon 

 her dress, burning her most severely. 



A lamp explosion which occurred last December in a van used 

 as a bedroom by an itinerant showman, at the so-called World's 

 Fair held at the Agricultural Hall, Islington, and which caused 

 the death of an infant, was of a somewhat different character to 

 the foregoing. The lamp, which was of the duplex-form and 

 was attached to a bracket, had been alight for some hours, when 

 a woman went, from a neighb luring van used as the dwelling 

 room, to extinguish it. She observed that while the lamp, or 

 only burning faintly, the oil in the reservoir was 

 alight. She placed her apron over the top of the chimney to 

 extinguish the lamp, when it at once appeared to explode, and 

 the burning oil set the interior of the van on fire. The woman 

 ran out for help, and a lad, protecting his head with his coat, 

 rushed in and brought out the infant which was lying upon the 

 bed, and which died from injuries received. The oil used in the 

 lamp was believed to be of high flashing point, being obtained 

 by the retailer who supplied it, from a firm dealing in a Scotch 

 shale oil manufactured by the Walkinshaw Company (known as 

 the " electric light " brand). A sample of the oil, as supplied by 

 the wholesale dealers, had a flashing point of 114 F., but a 

 portion of the oil actually purchased by the owner of the lamp 

 had a flashing-point of only 63 F. , and evidently consisted of a 

 mixture of the heavy oil and of benzoline. The oil in question 

 would naturally become exhausted of the volatile spirit after the 

 lamp had burned for some time, and the flame would then have 

 burned low in consequence of the heavy character of the residual 

 oil ; the lamp and its contents would have thus become highly 

 heated, and some accidental disturbance of the surrounding air 

 must have caused vapour generated from the heated oil and con- 

 tained in the air-space of the reservoir, to become inlla ned, the 

 oil itself being thereby ignited. By placing her apron hastily 

 upon the top of the chimney, the woman forced air into the 

 reservoir, and thus either caused a slight explosion to take place 

 or determined the breaking of the glass by the sudden change of 

 temperature. A lamp explosion, apparently due to the same 

 cause, occurred quite recently in the cabin of a smell steam- 

 launch on the Medway, near Chatham. 



: cases of undoubted lamp explosions, forlanately un- 

 attended by serious consequences, have come to the lecturer's 

 knowledge as having occurred in the billiard-rooms of barracks 

 where petroleum or paraffin oil was employed as an illuminant. 

 These lamps ate fixed over the billiard-tables, and generally 

 speaking the rooms have top- or sky-lights. In every instance 

 the lamp had been binning for several hours, and had probably 

 become more or less heated, especially as shades ot sheet tin 

 were placed over them as reflectors. In each case a portion of 

 the ylass reservoir was blown out by the expl ision, and the 

 oil, becoming ignited, burnt portions of the table on which it 

 fell. 



A careful investigation of accidents of which the foregoing are 

 illustrations, 1 together with a critical examination of the con- 

 struction of various lamps, and the results of many experiments 

 have, up to the present time, led the lecturer and Mr. Redwood 

 to arrive at several definite conclusions with respect to the im- 

 mediate causes of lamp-explosions and to certain circumstances 

 which may tend to favour the production of such explosions. 



If the lamp of which the reservoir is only partly full of oil be 

 carried, or rapidly moved from one place to another, so as to 

 agitate the liquid, a mixture of vapour and air may make its 

 escape from the lamp in close vicinity to the flame, and, by 

 becoming ignited, determine the explosion of the mixture exist- 

 ing in the reservoir. This escape may occur through the burner 

 itself, if the wick does not fit the holder properly, or through 

 openings which exist in s ime lamps in the metal work, close to 

 the burner, of sufficient size to allow flame to pass them readily. 

 A sudden cooling of the lamp, by its exposure to a draught or 

 by its being blown upon, may give rise to an inrush ot air, 

 increasing the explosive properties of the mixture of 

 vapour with a little air contained in the reservoir, and the flame 

 mp may at the same time be drawn or forced into the 



1 Mr. Alfred Spencer, of the Metropolitan B >ard of Works, has obligingly 

 furnished me with the official details of several of the accidents above referred 

 to— F. A A. 



