Page:Proceedings of the Royal Society of London Vol 69.djvu/126

118 discussed in full; in the former, those of analysis, unavoidable in this case, temperature, errors of filling pyknometers both with acid and water; in the latter, those of micrometer screws, divided circle, parallelism of quartz plates, as also the several effects likely to be produced by the various substances with which the acid

solutions of necessity came into contact. The results obtained by both methods are given in a series of tables and compared with those calculated from various equations for straight lines. These show that the physical properties display well-marked alterations at definite points in a word, such properties are discontinuous. Further, such points correspond very approximately to the concen-