The use of 7 3 y 0 sulphuric acid a t 41' for 1 hour or less caused the periodate lignin t o condense with 32y0 by weight of formaldehyde, a s shown by direct measurement and by employing C1a-formaldehyde. Analyses of the product, however, suggested that the elements of water had been expelled during the condensation and that the true yield was about 120%. Up t o 106% was recovered a s material insoluble in dilute sulphuric acid (Ross-Potter lignin), and about 9% as a more soluble form. The approximate equivalence in weight of the Ross-Potter lignin t o the original periodate (or Klason) lignin thus originated in the cancellation of three large errors. Formaldehyde periodate lignin failed t o yield vanillin when oxidized in alliali with nitrobenzene or cupric hydroxide. Oxidation with alkaline potassium permanganate, or with iodic acid a t not more than 200°, yielded benzenepentacarboxylic acid and perhaps mellitic acid (total, 3.9%) but none of the 1,2,4,5-tetracarboxylic acid recovered (together with the pentacarboxylic acid) from similar oxidations of periodate lignin. Convenient semimicro-scale oxidations with cupric hydroxide and with iodic acid, and the chromatography of the benzenepenta-and -tetra-carboxylic acids on paper, were described.
INTRODUCTIONI t is well kilown that the treatment of wood meal with 72y0 sulphuric acid near 20°, as in the I