The hydrodynamical theory of a Newtonian fluid flowing through a capillary viscometer is quite simple and yields a formula for the fluidity in terms of the measured efflux. In the case of non‐Newtonian fluids, on the other hand, the theory is not so simple. The usual practice among experimental rheologists is to assume some particular form of fluidity function, after which the corresponding efflux formula is obtained by an integration. If the efflux formula agrees with the experimental efflux data, the assumed fluidity function is thereby established as the correct one. However, it is sometimes difficult to find a fluidity function leading to an efflux formula which agrees satisfactorily with the data. In such cases the experimenter needs a more powerful method of analysis.