Page:Proceedings of the Royal Society of London Vol 69.djvu/103

Rh

It will be observed that the differences in the values of the constant are considerable, with the exception of the slight variation at 4 per cent., upon which it is not desired to lay stress ; further, the values of the constant are fairly simple integers.

The Descending Branch consists of five segments, (i) a rather flat portion from 54 to 63 per cent., (ii) a steep portion from 63 to 70 per cent., (iii) a more steep portion from 70 to 78 per cent., (iv) an abrupt descent from 78 to 95 per cent., and (v) a curved portion from 95 to 100 per cent.

The values for the constant a are summarised in the following table:

Percentage.

At 4.

At 14 -2.

At 24 -2.

v

5463

- 88

- 70

- 73

VI

63_70

-100

100

- 85

VII

70-77

-157

-150

-122

VIII....

7794

-205

-208

-195

It will be observed that the difference in value for the portions V and VI is not considerable.

As to Segment IX, or curved portion, it was thought at the earlier stage of the inquiry that the apparently irregular results obtained with acids of percentage strengths from 95 to 100 per cent, might arise from experimental errors, the possibility of which is greatly increased in the case of solutions difficult to manipulate and of high coefficient of expansion. With a view of deciding the matter, fresh samples of acids were prepared de novo after an interval of about fifteen months ; concordant results were obtained in the case of two samples of different past history and of approximately the same concentration, namely, 92'34 and 92 - 66 per cent, respectively, which pointed to an absence of any substantial error in the earlier work. Further determinations made with samples of concentration inter- mediate between those formerly made gave intermediate values.

The results are illustrated by certain curves, for a description of which the reader is referred to pp. 100-105.

Effect of Temperature, Though the mean straight lines representing the observations are not parallel with one another for the three temperatures, yet within the limits 4 24 - 2 there does not appear to be any great variation in the percentage, composition corresponding to the critical points as conditioned by temperature. It is very probable that at higher temperatures such variations would occur,