A previous communication2 (6) presented a systematic report upon the apparent viscosity of solutions of nitrocotton (12.10 per cent nitrogen) in a large number of solvents and solvent mixtures over periods of time up to one year. Many interesting and significant regularities appeared, of which three may be mentioned. (1) The best solvents give the least viscous solutions with the same nitrocotton, the viscosity being referred to that of the solvent itself in each case. (2) Mixtures are better solvents than pure liquids. (3) The apparent viscosity is largely due to a structural effect,3 since it is altered by shaking, change of solvent, previous treatment, previous precipitation and recovery of the nitrocotton, and because of its enormous magnitude. All this is additional to any effect of the length of the molecule itself, so much stressed in subsequent years by Staudinger and others. The effects quoted above cannot reasonably be ascribed to alteration in the length of the molecule itself, more especially since changes occur in both directions.Many questions demanding further study arose, some of which are dealt with in the present report .
EXPERIMENTAL METHODSThe nitrocotton "1919" has been described, together with the method of drying, storage, and preparation of the solutions (6). Units of weight, methane" should be "phenyl urethane" and "ortho tolyl methane" should be "benzyl phenyl urethane"; p. 330, table XI, the viscosity of pure ethyl phthalate (Nobel's; December, 1923) should be 0.04215 instead of 0.4215; p. 333, third line from the bottom, "less impure" should be "less pure."3 This was developed as a general explanation of the high viscosity of numerous colloids by J. W. McBain (J. Phys. Chem. 30, 239 (1926)).