302 PHILOSOPHICAL TRANSACTIONS. [aNNO I766. 



might perceive the flame to spread gradually through the bottle. In the next 3 

 trials, though they made an explosion, yet I could not perceive any light within 

 the bottle. In all probability, the flame spread so instantly through the bottle, 

 and was so soon over, that it had not time to make any impression on my eye. 

 In the last mentioned trial, namely, that with equal quantities of inflammable 

 and common air, a light was seen in the bottle, but which quickly ceased. 



With 6 parts of inflammable to 4 of common air, the sound was not very 

 loud: the mixture continued burning a short time in the bottle after the sound 

 was over. With 7 parts of inflammable to 3 of common air, there was a very 

 gentle bounce or rather puff": it continued burning for some seconds in the belly 

 of the bottle. A mixture of S parts of inflammable to 2 of common air caught 

 fire on applying the flame, but without any noise: it continued burning for some 

 time in the neck of the bottle, and then went out, without the flame ever ex- 

 tending into the belly of the bottle. It appears from these experiments, that this 

 air, like other inflammable substances, cannot bum without the assistance of 

 common air. It seems too that, unless the mixture contains more common than 

 inflammable air, the common air therein is not sufficient to consume the whole 

 of the inflammable air; by which part of the inflammable air remains, and bums 

 by means of the common air, which rushes into the bottle after the explosion. 

 In order to find whether there was any difference in point of inflammability 

 between the air produced from different metals by different acids, 5 different sorts 

 . of air, namely, 1 . Some produced from zinc by diluted oil of vitriol, and which 

 had been kept about a fortnight; 2. Some of the same kind of air fresh made; 

 3. Air produced from zinc by spirit of salt ; 4. Air from iron by the vitriolic acid; 

 5. Air from tin by spirit of salt; were each mixed separately with common air in 

 the proportion of 2 parts of inflammable air to 7-^ of common air, and their 

 inflammability tried in the same bottle that was used for the former experiment, 

 and with the same precautions. They each went off" with a pretty loud noise, 

 and without any diflference in the sound that I could be sure of. Some more of 

 each of the above parcels of air were then mixed with common air, in the pro- 

 portion of 7 parts of inflammable air to 34- of common air, and tried in the 

 same way as before. Each of them went off" with a gentle bounce, and burnt 

 some time in the bottle, without my being able to perceive any difference be- 

 tween them. 



To avoid being hurt, in case the bottle should burst by the explosion, I have 

 commonly, in making these sort of experiments, made use of an apparatus con- 

 trived in such manner, that by pulling a string, I drew the flame of a lamp over 

 the mouth of the bottle, and at the same time pulled off^ the cap, while I stood 

 out of the reach of danger. I believe however that this precaution is not very 

 necessary, as I have never known a bottle to burst in any of the trials I have made. 



