IS THE VISCOSITY A DEFINITE PHYSICAL QUANTITY? 61 <[

I'

has also been taken as the standard and equal to unity. Still \\

other standards have been employed. The principal advantages I

in this form of expression are the saving of labor in calculation, ¥f

the avoidance of inconveniently small fractions, and the use of a f^

common liquid as standard. This advantage however is more I/1

than offset by the disadvantages. The proper corrections are \f
never applied to specific viscosities and consequently the values ;;'

are not really comparable among themselves. Certainly they are ;)

inconvenient to use for reference and for comparison with vis-
cosities calculated in other ways. The time necessary for the

preparation of the substances is nearly always great enough to ^

justify the inconsiderable expenditure of time necessary for the fi

proper reduction of the data to absolute units. ' '

Of course much depends upon the disposition of the apparatus -^

used in the measurement. Some forms of apparatus will not \'<*„

permit accurate estimations of the viscosity to be made. But £f

given an apparatus which is well-suited for precise measurements, |[

the time required for making a measurement is no greater, and |j|

may be much less, than in the less accurate forms of apparatus. !tf i

The Centipoise.—In expressing viscosities, it is possible to p

secure simultaneously the advantage of expression in absolute f |

units with the advantages of viscosities relative to some common (,f

substance as standard. It is proposed to name the absolute unit J'|

of viscosity after Poiseuille the "poise," and consequently the ;f|

subnrultiple of this unit which is one-hundredth as large the |f

'' centipoise " (cp). It so happens that one centipoise is almost ex- ; J
actly the viscosity of water at 20°C, hence absolute viscosities
expressed in centiposes are also specific viscosities referred to water
20°C as standard. To be sure the viscosity of water is not ex-
actly one centipoise at 20° C but it is 1.005 which is unity within
the limits of possible experimental error in ordinary measurement
(cf. Appendix D, Table II).

All fluidities are expressed in absolute units, water at 20°
having of course a fluidity of 100 units.

4
r