350 



SCIENCE. 



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



holds that the relative volatility of petro- 

 leum oil is a subject which is not sufficiently 

 known and appreciated. By comparing the 

 loss of weight during 24 hours of oils ex- 

 posed in shallow vessels under similar con- 

 ditions, he found at 66° F. an American 

 water white oil of 106° flash point lost 20.4 

 per cent., an oil of 75° flash point lost 27.4 

 per cent., and a Russian oil of 84° flash 

 point lost 28 per cent. 



In observations that I have made it was 

 very apparent that the form and material 

 of the containing vessel are most important 

 factors in these volatilization experiments. 

 I have found, for instance, that a given vol- 

 ume of gasoline placed in an uncorked vial 

 and exposed to the ordinary atmospheric 

 conditions of a laboratory required 10 weeks 

 for complete volatilization when the same 

 volume of the same lot of gasoline placed 

 in an evaporating dish standing beside the 

 bottle volatilized completely in 8 hours. 

 The rate of evaporation of various hydro- 

 carbons under the same conditions has been 

 studied by Boverton Redwood.* 



A menace in the use, storage and trans- 

 portation of these liquids rests in the 

 rapidity with which their vapors diffuse 

 through the air and form an explosive train 

 which reaching out to a source of ignition 

 flashes back with extreme i-apidity through 

 the entire train and to its point of origin. 

 Sir Frederick Abel cited an instance of thisf 

 which happened at the Royal College of 

 Chemistry in 1847 when a glass vessel in 

 which benzene was being converted into 

 nitro-benzine broke and allowed the warm 

 liquid to escape and flow over a large sur- 

 face. Though the apartment was 38 feet 

 long, 30 feet wide and 10 feet high, and the 

 only ignited gas jet was at the end of the 

 room most remote from the glass vessel, yet, 

 in a very brief space of time after the vessel 



* ' Detection of Inflammable Gas and Vapor in the 

 Air,' Frank Clowes, p. 191, 1896 ; London. 

 t Roy. Inst, of Great Britain, March 13, 1885. 



broke, a sheet of flame flashed from the gas 

 jet and traveled along the upper part of the 

 room to the point where the fluid lay scat- 

 tered. 



Also he cites the explosion of benzoline 

 at the mineral oil store in Exeter in 1882. 

 The store rooms were arched caves in the 

 side of a bank facing a canal and separated 

 from it by a roadway about 50 feet wide. 

 There was a standing rule forbidding any 

 Wght being taken to any of these store 

 rooms when they contained petroleum 

 spirit, but on the day in question it was 

 desired to remove some of the benzoline in 

 the early morning and the foreman visited 

 the store rooms before daylight to make 

 ready for the work. Forgetful of the rule, 

 he carried a lighted lantern, which he placed 

 on the ground some 27 feet away from the 

 cave, and was proceeding to open the door 

 when he observed a strong odor of benzo- 

 line and almost immediately noticed a flash 

 of flame proceed from the lantern to the 

 store and had barely time to turn to escape 

 when an explosion took place which blew 

 the doors and lantern across the canal and 

 inflamed the spirits in the store rooms. 



Of course, the distance that these vapors 

 will travel will be determined by the cir- 

 cumstances of each individual case, but in 

 the case of the fire at the L. & N. "W. R. 

 R. Co.'s gas factory in February, 1897, 

 through which the hydrocarbons in a cyl- 

 inder that was being rolled across the yard 

 about the works became ignited, the nearest 

 source of ignition was found in the boiler 

 fires, which were 60 feet away.* 



Conditions such as these are more likely 

 still to obtain when these inflammable and 

 volatile substances are stored in enclosed 

 spaces, such as the hold of a vessel during 

 transportation , and they have been the cause, 

 under these conditions, of many fright- 

 ful accidents. As an example of these we 

 have the case of the explosion on Novem- 



* Kept. H. M. Insp. Exp. 22, 57; 1898. 



