The mistake is often made of identifying Labarraque's liquor and even commercial eau de Javel with Dakin's solution. But Daufresne has shown by simple methods, that, from the biological as well as the chemical point of view, these three solutions behave in very different ways.

Amongst the reactions which may be cited for this purpose, two are particularly characteristic. The phenol-phtalein reaction and the effect upon skin. In these experiments the three solutions are brought to a convenient strength of (O gr. 50 %, Fr.) of hypochlorite of soda. The action upon skin has already been described (p. 23). We shall here give only the action upon phenol-phtalein.

If 20 c.c. of the solution to be examined are poured into a beaker and on the surface are placed a few centigrammes of phenol-phtalein in powder, it is seen that:-

1st. Eau de Javel and Labarraque's liquor immediately colour the particles of phenol-phtalein an intense red, and the slightest shaking will suffice to communicate to the whole of the liquid a bright red colour, which slowly disappears under the decolorising action of the hypochlorite.

2nd. Dakin's solution, under the same conditions, does not give any colour to the particles of phtalein, and it is only after vigorous and prolonged shaking that the liquid becomes of a faint rose tint.

Then, if one seeks the amount of alkalinity which a solution must possess in order to give so much colour to powdered phtalein, it is found that only solutions containing at least 0.2 per cent. of caustic alkali will give to the phtalein test a similar degree of colour. Carbonate of soda gives only an almost imperceptible tinge to the particles of phtalein, and a rosy tint to the liquid: that same solution gives no colour if it only contain 20 per cent. carbonate of soda.

Therefore Labarraque's liquor and eau de Javel each contain a small quantity of caustic soda, revealed by the phenol-phtalein test, and which might readily be foreseen after examination of their mode of preparation.

In fact, Labarraque's liquor and many samples of commercial eau de Javel are obtained by double decomposition of a solution of chloride of lime and a solution of carbonate of soda. All the constituents of chloride of lime (hypochlorite of calcium, chloride of calcium, slaked lime) are able to react upon carbonate of soda, giving respectively hypochlorite of soda, chloride of sodium and caustic soda. This caustic alkali, which constitutes the irritating element most to be dreaded in hypochlorite solution, certainly exists in the earlier stages of preparation as given by Dakin; but later it is neutralised by excess of boric acid. We have seen why it is not formed in the process described by Daufresne.