ing to Tamman, ice can exist in no less than six forms, depending on the conditions of pressure and temperature. No substance can be studied better from the electronic standpoint.
On the ground of analogy, the chemist can foresee the exist- ence of an active Hydronol and of a variety of neutral Hydrones corresponding to the paraffinoid polymethylenes, thus
H 2 O H 2 O.OHj H 2 O<
OH OH 2 H 2 O.OH,
No valid method of determining the complexity of the mole- cules either of water or of ice is known; all that can be asserted is, that water especially must be a mixture and that its compo- sition not only may but must be subject to considerable varia- tion as the temperature is changed or substances are dissolved in it. Whatever the composition of the mass, at the surface, the simple molecules of hydrone (OH 2 ) must be present in maxi- mum proportion;'and this will also be the condition at the sur- face of solid particles suspended in an aqueous solution.
As the most active solvent in water must be the hydrone mole- cules, in a solution in which fine particles are suspended the liquid layer at the fluid-solid interface should be more concentrated than the general body of the solution. Hence the special activity of enzymes and other particulate agents: apart from any special attractive influence exercised by the solid surface, the layer at the interface is likely to be specially active as a solvent.
The enzymes, however, exercise a selective activity which is altogether peculiar each enzyme can induce the hydrolysis, if not of a single compound, at most of a set of structurally similar compounds. Thus the enzyme urease will act only on urea. Invertase acts only on cane-sugar or derivatives of this sugar in which its special structure is retained and only an addition made to the molecule.
In the case of glucose, a large number of compounds are formed by the introduction, in place of either the one or the other of two terminal hydrogen atoms, of some equivalent group. Two series of glucosides are thus produced, known respectively as a- and /3-glucosides. An enzyme is present in yeast (maltase) which will induce hydrolysis of all the a-glucosides; the bitter- almond contains an enzyme which acts only on the /8-compounds.
The only possible explanation of this behaviour seems to be that the enzyme is structurally related to the compound which it affects, so that it actually fits upon it and grasps it, as it were. This view involves the further assumption that the specific agent of change is also carried by the enzyme, as the amount of acid which suffices to determine the exercise by the enzyme of its maximum activity is so small that it scarcely seems prob- able that the rapid action is the consequence of the mere con- centration of this acid together with the hydrolyte at the surface of the enzyme; nor is such an assumption compatible with the selective activity of enzymes. More probable is it that an acid radicle is operative which the enzyme itself carries, this being in such a position that it is brought into proximity with the attached molecule (the hydrolyte) at the point at which hydrol- ysis takes place the acid which is added serving to maintain this radicle free and also acting as the necessary electrolyte, as in the case of platinum.
It should be added that, although platinum and similar catalysts are not structurally selective agents, their action is in some respects limited as, however, is that of most chemical agents. Thus, whilst hydrogen and oxygen interact at a platinum surface, a mixture of carbonic oxide with oxygen remains un- affected; indeed, carbonic oxide interferes with the oxidation of hydrogen in presence of platinum; this behaviour, however, is perhaps less a consequence of the lack of affinity of carbonic oxide for platinum than of intrinsic peculiarities of the gas. Platinum and similar catalysts, especially nickel, have a very wide range of activity as hydrogenising agents.
The question remains how is the action of platinum effected; is it merely a physical condensing agent or is it to be regarded as acting chemically? The view has long been held that it may combine with oxygen if not with hydrogen and that it pro- motes oxidation through the intermediate formation of an oxide.
Recently, Willstatter has shown that platinum and palladium are without action even as hydrogenising agents if they have been freed from oxygen. He advocates the view that a compound is formed, which is both hydride and peroxide, in which hydrogen is present in a more readily dissociable form than in the hydrides of the metals. Spongy platinum, prepared by reducing chloro- platinic acid with formaldehyde in presence of caustic potash, may be deprived of oxygen by suspending it in glacial acetic acid and passing hydrogen through the liquid, either in the course of 30 hours at atmospheric temperature or in 8 hours at 50 to 60. Such a product is insoluble in muriatic acid and does not liberate iodine from an acid solution of potassium iodide; it will not condition the hydrogenation of benzene to hexamethylene, etc., but acquires the power when shaken, during a short period, with air. During hydrogenation, the catalyst invariably becomes deoxidized by the action of the hydrogen and needs revivification by oxygen. Willstatter suggests that the metal is converted
into either the peroxide Pt< o or the corresponding perhydrol
and that this is convertible into the hydride
II ,
li-
O-OH OH
By this assumption, the activity of platinum in promoting oxidation is brought entirely into line with that of ferrous sul- phate, which, it has been pointed out, is probably active as a perhydrol. Interesting light is also thrown on the hitherto enigmatic behaviour of haemoglobin, which combines directly with oxygen, forming oxyhaemoglobin, by the fact that the oxidized platinum catalyst may be deprived of its oxygen and rendered nearly inactive by continuous exhaustion with a high vacuum pump, the means by which oxyhaemoglobin may be entirely deprived of oxygen: the parallel is made all the more remarkable by the fact that the oxygen may be displaced from oxyhaemoglobin by carbonic oxide, which renders platinum inert towards hydrogen. Oxyhaemoglobin has a big molecule, as it is composed of a protein in association with haematin ; if the oxygen be present in it, perhaps in combination with the iron atom, in the form of perhydrol and it acts in the blood corpuscle as a particulate agent, the remarkable oxidizing power of the blood may be reckoned among the actions promoted by catalysts.
It is customary to regard haemoglobin as a " colloid," but in using this term we are again in difficulty owing to the lack of a clear definition of its precise connotation. Latterly the word has been used almost as the synonym of the state of very fine sub-division any substance present in suspension in a liquid in a very finely divided state has been spoken of as a colloid.
Originally this was not the meaning associated with the term by Graham, who introduced it and applied it generally to glue- like substances. He appears to have thought of the colloid as soluble but as merely opposite in the scale of solubility to the ordinary crystalline, more or less easily soluble substances of relatively low molecular weight in fact, as a big, lumbering molecule, with slight affinity for the solvent and therefore ready to separate from it in the pectous or particulate form. Unfor- tunately, not only has the connotation of the term been altered but a confused language has grown up about the term which renders the consideration of the activity of substances in the particular state specially difficult. Far worse, the attempt has been made to constitute so-called Colloid Chemistry a separate discipline, the designation being arbitrarily confined to sus- pensions of fine particles varying from one thousandth (/*) to one millionth (/x/u) of a millimetre in diameter.
If this definition be accepted, the " colloids," when separated, in the particulate state, from solutions should function as cata- lysts under favourable conditions; and this appears likely to be true. One of the few cases apart from the action of enzymes which are selective catalysts of a colloid having been shown to act specifically as a catalyst is in the production of hydrazine from ammonia, by the action of a hypochlorite, by Raschig's method, an interaction which is promoted by the addition of glue,
