The method of Schickh, Bin2 and Sclzidz for the iodination of 3-hydroxypyridine t o give a hyclroxymonoiodopyridine is also suitable for iodinating 2-and 4-hydroxypyridine ; 2-(or 4.) hydroxy-3,5-diiodopyridine is produced as a by-product from 2-and 4-hydroxypyridine, respectively.Above -255" 4-hydroxy-3-iodopyridine is converted into a mixture containing 4-hydroxy-3,5-diiodopyridine. IntroductionDirect halogenation of 2-hydroxypyridine in weakly alkaline, polar solvcnts, readily leads, as is well known, to polychloro-and polybromo-compounds, while on iodination, 2-hydroxy-5-iodopyridine is produced in poor yield 1.With 3-hydroxypyridine direct halogenation with bromine in pyridine affords 2-bromo-3-hydroxypyridine 2 ; nothing is known on chlorination, while iodination in weakly alkaline, aqueous media affords 3-hydroxy-2-iodopyridine in yields of 65-90:,

No abstract

Measurements are described on the e.mJf. of cells with electrodes of paraffin and polystyrene membranes of different composition in a number of solutions contdning uni-or higher valenced cations. From these researches it follows &at uni-and higher valenced cations behave antagonistically.In our first paper of this series 1 we stated: "The different behaviour of uni-and bi-valent ions is, however, not clear".This statement concerned the observations made in experiments described in that paper. Paraffin membranes more or less specific for potassium or calcium ions show an antagonistic effect for these ions. The effect, however, was not the subject of the results described at that time.Continuing our researches we proved that paraffin membranes con-

Kinetic experiments indicate that the first step of the oxidation of phloroglucinol by Fe (CN) a ' is reversible. The equilibrium concerned, however, is not stable, since one of the species participating, i.e. the first oxidation product of phloroglucinol, is broken down irreversibly. Some data are given about the reaction rates concerned, and the existence of a free radical 1) H.

In previous papers 1) we described researches on the potentiometric titration of relatively strong acids, adsorbed on activated charcoal. From the curves we concluded, that the strength of the adsorbed acid had decreased and that the systems containing activated charcoal have a marked buffer-action. There is no question of a suspension effect; the pH reading is taken up in the suspension.In this paper we shall develop a theory about the properties of the adsorbed acid molecules, a theory, which makes it possible to calculate the titration curves, though not completely, and which is similar to calculations of titration curves of solutions of macromolecules. W e imagine a particle of the activated charcoal on which the molecules of a dissolved acid have been adsorbed, as a poly-acid of unknown basicity. The adsorption forces on the surface of the activated charcoal particle will differ from spot to spot. As these adsorption forces influence the properties of the adsorbed molecules, we may expect that the dissociation of the hydrogen-ions will be influenced differently. Molecules, which are adsorbed strongly will show a more difficult dissociation of their hydrogen-ions. in other words they will behave as a weaker acid than molecules which are less strongly adsorbed.In accordance with these views we found that it is was impossible to describe titration curves of adsorbed acids by formula ( 5 ) .Our curves are more comparable with those of a mixture of acids, each one having its own dissociation constant. Yet, these dissociation constants will not be markedly different, as is the case in a mixture of several distinct components: they will change gradually.The theory of the titration constants of multivalent substances has