We study several methods of describing `explicit' solutions to equations of Korteweg--de Vries type: (i) the method of algebraic geometry (Krichever, I.M. Usp. mat. Nauk 32, 183-208 (1977)); (ii) the Grassmannian formalism of the Kyoto school (iii) acting on the trivial solution by the `group of dressing transformations' (Zakharov, V. E. & Shabat, A. B. Funct. Anal. Appl. 13 (3), 13-22 (1979)). I show that the three methods are more or less equivalent, and in particular that the `$\tau $-functions' of method (ii) arise very naturally in the context of method (iii).