ts TEMPERATURĘ C0EFFICJEN7
mtduMism will independently lead to the same Muation. The important feature is tftat \J, m Dunn's form of the erjuation, has the dimensions of an energy. Curves plotting log rj against r/T are fairly linę ar for most fluids. With appar ently solid tmterials for which viscosities can be measured, thcsc curves are not realiy iinear. t&upercoolad liquid% such as giasses and shellacs, give fairly stmight linęs over a moderate temperaturę rangę.
Leon'teva compares the Andrade formula with a rather morę compJex formula, proposed by Khalkin, in the case of metaphosphates and borates. Therc is little to choose between the formula: with regard to their agreement with the data.
Lederer calculates that
where a is an energy related to molecular weight and $ is the entropy. Since he says that this eguation cannot be integrated to a finite form, he concludes that no ńnite deńnition of the viscosity-temperature coefficient is theoreticaUy possible. This cannot be said to contradict the conclusions of Sheppard and others, sińce they make assumptions I in deriving their eguations which cannot be more than approximations to the trnth. A similar con-clusion will later be drawn with regard to the definition of plasticity.
There are a few materials which have »very anomalous viscosity-temperature coefficients, of I which perhaps the most striking is sulphur. At 140° C. sulphur has a viscosity of about go c.g.s. i units (or " poises," as they are called, the reciprocal I poise being called the " rhe"). On heating to I 1800 C. this rises to 500 poises, and then falls j
19
CHEMICAL CONSTITUTION
gradually on further heating until, at 4009 C., it has fallen to 20 poises again, The reasons for this are not yet lully nnderstood (vide Farr and McLeod).
The way in which viscosities vary with the number of carbon atoms in a homologous series of organie compounds has been extensively investigated. McLeod (vide Hatschek's " Viscosity of Liquids ") finds correlations between viscositics and coefficients of expansion. With an inerease in the number of carbon atoms the viscosity inereases as the coefficients of expansion decrease. He calculates with some success the viscosities for a series of hydrocarbons, acids, alkyl halides, ketones, alcohols, esters, etc.
Bingham suggests relationships between fiuidity, temperaturę and molecular weight; and Porter, studying liąuids in pairs, finds that if tx and łt are chosen as a pair of temperatures for two liąuids, 1 and 2, at which their yiscosities are the same, then curves plotting txlt2 against tx are fairly Iinear. This conclusion is justified for very varied pairs of liąuids.
Most of the work that has been done on Chemical constitution in relation to viscosity has had to do with the homologous series. Hatschek, in his book, deals very fully with it, ąuoting from Gartenmeister. The work is most interesting, but has little bearing on industrial rheology, and cannot, therefore, be dealt with in any detail here. The conclusions arrived at are briefly two; first, that for different compounds at different temperatures, having the same number of carbon atoms, viscosity divided by molecular weight is constant. Examples of this are propyl and allyl chlorides, bromides and iodides. Secondly, that the viscosity divided by the sąuare of the molecular weight is constant at all tempera-