The Atomic Arrangement in Glass

drous Rochelle salt, and the third we tentatively suggest corresponds to solid complex. At all events, the tie-lines show no indication of running to the sodium hydroxide comer. For the reason previously given, it cannot be expected that the tie-lines should intersect in a point. The appearance of the solid phase a t this stage is remarkable. It swells in bulk so as to occupy almost the total volume and resembles a stiff gel. The numerical results are contained in Table 111. The triangular diagram is interesting and has theref ore been reproduced. It is hoped shortly to communicate similar work which has been done on Work is now proceeding camphoric acid and substituted tartaric acid. on the acid and alkaline racemization of mandelic acid.