78 Intelligence and Miscellaneous Articles, 



Were p of the form \l^(p) -^ x(v) , the quantity added to q would 

 =\^(t;), and would vanish after the restoration to the original volume. 

 In general, however, p would produce a change in the quantity of 

 heat, which evidently would be always small, inasmuch as a sudden 

 or quick change in p is very improbable. — PoggendorfF's Annaletiy 

 xcvii. p. 30, January 1856 ; and Silliman's Journal for May 1856. 



EXPLOSIVE ACTION OF SODIUM ON WATER. 

 BY FREDERICK W. GRIFFIN, PH.D. 



The action of sodium on water contained in a glass vessel is such 

 an elegant proof of the composition of that liquid and other points, 

 as to lead to its being occasionally introduced in lectures. I have, 

 however, found the experiment liable to a grave accident, which 

 ought altogether to banish it from public demonstrations. Some 

 years ago, during my private course of lectures, I passed a piece of 

 the metal, about a quarter of an inch square, into a tube filled with 

 water, 18 inches in length by 1 in diameter. When the action was 

 nearly over, a powerful explosion occurred, which forced the tube 

 (weighing upwards of a pound) violently through my hand and 

 dashed it to pieces against the ceiling. As I had often before per- 

 formed the experiment with perfect safety, I presumed that air must 

 have somehow got mixed with the hydrogen ; the more so as, simply 

 holding the sodium in the fingers, I had, to slip it under more 

 quickly, brought the mouth of the tube very near the surface of the 

 water. On all subsequent occasions I placed the sodium in a little 

 tube closed at one end, which it almost filled, and stopping the mouth 

 with the finger, opened it below the larger tube, which was kept at 

 least a couple of inches under water, so that there was no possibility 

 of letting in air by any sudden jerk or otherwise. All went off well 

 for several times, till at a public lecture in Devonshire an explosion 

 resulted more violent than the first, and the tube was blown into 

 splinters which strewed the floor of one half of the room, and slightly 

 wounded several persons. Since that occurrence I have relinquished 

 showing this experiment in public at all, though numerous trials 

 appear to prove that a piece less than a pea may be used with 

 safety, though there is sometimes a slight concussion at the end. 

 'Vhe cause of the detonation remains to be explained. In the last 

 instance at any rate, it is quite certain that no explosive mixture with 

 air was formed, and I have little doubt that the effect proceeds from 

 the water round the sodium being thrown into the spheroidal state. 

 This view seems confirmed by the fact, that at the first moment of 

 contact a large quantity of gas is always liberated, but the action 

 speedily becomes weaker, and the evolution of hydrogen extremely 

 slow. In all probability the metal is then merely decomposing 

 the atmosphere of aqueous vapour around it ; and when the piece 

 is small, it disappears tranquilly in this way ; when it is larger, 

 so that the action is prolonged, its temperature slightly falls, contact 

 ensues, and a burst of gas and steam takes place with explosive 

 violence. In both cases the tube waa three-quarters full of gas» and 



