1885.] on Accidental Explosions by Non-exjplosioe Liquids. 239 



air existing in tlie lamp may be more violently explosive, because the 

 proi)ortion of the former to the latter is likely to be lower and nearer 

 that demanded for the production of a powerfully explosive mixture. 

 If the quantity of oil in the lamp reservoir be but small, and the 

 air-space consequently large, the ignition of an exjilosive mixture 

 produced within the lamp will obviously exert more violent effects 

 than if there be only space for a small quantity of vapour and air, 

 because of the lamp being comparatively full. If the wick be lowered 

 very much, or if for some other reason the flame becomes very low, 

 so that it is burning beneath the metal work which surrounds and 

 projects over the wick-holder, the lamp will become much heated at 

 those parts, and the tendency to the production of an explosive 

 mixture within the space of the lamp will be increased, while, at the 

 same time, heat will be transmitted to the glass, and it will be corre- 

 spondingly more suscejDtible to the effects described as being exerted 

 by its sudden exposure to a draught. Experiments have demon- 

 strated that a lamp containing an oil of high flashing point is more 

 liable to become heated than a comparatively light and volatile oil, 

 in consequence of the much higher temperature developed by the 

 combustion, and of the comparative slowness with which the heavy oil 

 is conveyed by the wick to the flame. It therefore follows that safety 

 in the use of mineral oil lamps is not to be secured simply by the 

 employment of oils of very high flashing point (or low volatility), 

 and that the use of very heavy oils may even give rise to dangers 

 which are small, if not entirely absent, with oils of comparatively low 

 flashing points. The occurrence of such an accident as that in the 

 training-ship Goliath, already referred to, which was brought about 

 by a boy letting fall a lamp which had been alight all night, and which 

 was so hot that he could no longer hold it, appears to be primarily 

 ascribable to the use of an oil of very high flashing point ; and the 

 accident at the Agricultural Hall furnished another illustration of the 

 kind of danger attending the use of such an oil. 



The character of the wick very materially afiects not only the 

 burning quality of the lamp, but also its safety. A loosely plaited 

 wick of long staple cotton draws up the oil to the flame regularly 

 and freely, and so long as the oil be not very heavy or of very higji 

 flashing point, and therefore difticultly volatisable or convertible 

 into vapour (by so-called destructive distillation), the flame will con- 

 tinue to burn brightly and uniformly, with but little charring 

 effect upon the wick ; that is to say, the extremity of the latter will 

 only be darkened and eventually charred to a distance of much 

 less than a quarter of an inch downwards, and it will not be until 

 the partial exhaustion of the oil-sui3j)ly diminishes the size of the 

 flame and induces the user to raise the wick, that the latter will 

 become more considerably charred. But, if the wick be very tightly 

 plaited, and made, as is not unfrequently the case, of a short staple 

 cotton of inferior cajDillary power, the oil will be less copiously 

 drawn up to the flame ; as a consequence, the length of exposed wick 



