In Fig. 2, we have plotted AH against p1I2 at 0, 20, 40 and 60'. We have also indicated the limiting slope of the Debye and Huckel theory for the relative partial molal heat content of the ions of water. Our results should approach these slopes in very dilute solution (m = 0.001 or less). Since they cannot be extended to low dilutions with a high accuracy, no attempt has been made to employ the limiting laws for the purpose of extrapolation. As a confirmation of the correct character of the results a t higher concentrations, we may compare our data with values of the same quantity determined from calorimetric measurements of the heat of neutralization of hydrochloric acid and sodium hydroxide by Richards and Roweg at 20'. Their results are represented by (9) Richards and Rowe, THIS JOURNAL, 44, 684 (1922), the dots in Fig. 2. limit of 40 cal. The agreement is within the summary 1. Measurements of the cells H 2 I NaOH (0.01), NaCl (m) I AgCl I Ag Ht I HC1 (0.01), NaCl (m) I AgCl I Ag in aqueous solution from 0 to 60' have been made.From these the value of the ionic activity coefficient product of water in sodium chloride solutions has been computed.It is shown that the logarithm of the ionization in alkaline halide solutious, log mHmOH, at a given concentration varies nearly linearly with the sum of the reciprocal of the ionic radii. This is only a rough approximation but serves to show that ions of smaller radii cause the greater dissociation of the solvent.The total heat content of the ionization of water and the relative partial molal heat content of the hydrogen and hydroxyl ions in sodium chloride solutions has been determined from 0 to 60' . Good agreement with the calorimetric heat of neutralization is found a t 20". NEW HAVEN, CONNECTICUT RECEIVED JUNE 18, 1935 2.
3.
4.The chemical investigation of vitamin B1 was undertaken in this Laboratory on the invitation of Mr. R. R. Williams, who acquainted us with the sulfite cleavage reaction1 and kindly furnished a supply of the crystalline vitamin and its basic cleavage product. Early in the course of our work we were able to suggest the presence of a thiazole nucleus in vitamin BI, and this suggestion was incorporated by Williams in his tentative formula.2In view of the presence of sulfur in the vitamin molecule, our attention was primarily directed toward elucidation of the manner in which this element is combined, by the use of methods recently developed4 for the study of sulfur in proteins. In the preliminary stages of this investiga-(1) Williams, Waterman, Keresztesy and Buchman, THIS JOUR-(2) Williams, ibid., 67, 229 (1936). (3) Windaus, Tschesche, Ruhkopf, Laquu and Schultz, Z.