and on the Sulphur ets of Lime, 16 J 



the consideration that his first experiment would indicate 

 22*9 oxygen per cent, in air, which cannot be allowed ; and 

 his last experiment that 8*8 oxygen must have combined with 

 20*6 hydrogen instead of 17*6, which is equally inadmissible. 



Since the period 1784? it has been found by various che- 

 mists that in mixtures of oxygen and hydrogen, as well as in 

 other similar ones, the electric spark does not always cause 

 an explosion, and when it does a complete combination does 

 not always take place, but that in the residue sometimes por- 

 tions of both the ingredients may be found. The limitations 

 and restrictions are now pretty generally known ; and with 

 regard to the mixtures of common air and hydrogen, I pub- 

 lished a letter in the 10th volume of the Annals of Philosophy, 

 (New Series) page 304, in which I showed the limitations 

 found by my own experience to be as under : 



Common air and hydrogen in which the oxygen is only yVth, 

 or from six to seven per cent, of the whole mixture, do not 

 explode. 



Common air and hydrogen in which the oxygen is only y^th, 

 or seven per cent, explode imperfectly, leaving both oxygen 

 and hydrogen. 



Common air and hydrogen in which the oxygen is from 

 -j^^^th to T^.^T-th, or from eight to fourteen or fifteen per cent., 

 fire leaving hydrogen and azote only. 



Common air and hydrogen in which the hydrogen is ^^Vu^^ 

 to |th, or from fourteen to thirty per cent., fire and leave 

 oxygen and azote only. 



Common air and hydrogen in which the hydrogen is Jth to 

 y^^th, or from eight to twelve per cent., fire imperfectly, and 

 leave oxygen, hydrogen, and azote. 



Common air and hydrogen in which the hydrogen is y^jth or 

 less than seven per cent., do not explode. 



It should be observed that when one of the gases is so far 

 deficient as not to allow of an explosion by a single spark, the 

 effect may be obtained by a current of sparks for a longer or 

 shorter period, accompanied by the requisite diminution of 

 volume. In such instances where the effect is produced only 

 by A current of sparks it may be proper here to suggest the 

 reason. When mixtures explode perfectly but feebly, we see 

 the flame, lighted by the spark, to run down the eudiometer 

 till it reaches the water; when they explode still more feebly, 

 the flame runs perhaps half-way down the tube and is extin- 

 guished before it reaches the water. There scarcely can be a 

 doubt that the extinction must be occasioned by the cooling 

 effect of the eudiometer and of the intermixture of the mass 

 of air which has to be heated by the feeble flame. Another 



Phil. Ma^r, S. 3. Vol. 12. No. 73. Feb, 1838. X 



