cent, acid, or approximately of the composition HN0 3 .2H 2 ; further, Pickering (loc. cit.) has obtained crystals of this composition by freezing. As regards the contractions, the change at 70 per cent, appears to be more marked than that at 63 per cent., and it might be a matter of judgment if the results from 54 to 70 per cent, might not equally well be represented as upon one curve and not upon two straight lines. The latter view has been adopted to bring the results obtained within these limits into uniformity with the rest.
As regards the abnormality at 95 to 100 per cent., it is probable this is the result of a wholly different cause ; a similar phenomenon was observed in the electric conductivity, and the suggestion was put forward of an initial decomposition into water and nitric anhydride, but since the latter, when passed into nitric acid of about 100 per cent, concentration, forms a substance of composition 2HN03.NoO r , or 2N 2 5 .HoO (a colourless liquid of specific gravity 1*642 at 18, crystal- lisable at - 5), both the contraction and the electric conductivity numbers might be expressed in terms of this substance, and not in terms of nitric acid 2HN0 3 or N^0 5 .HoO. At present the physical constants of the crystallisable compound have not been sufficiently examined.
IX. The Degree of Discontinuity.
Though in the diagrams of curves the mean straight lines are pro- duced until they intersect, yet the object thereof is to determine each separate origin of co-ordinates; it is not ^intended to represent an irregularity occurring suddenly or an abrupt change of events. The results in themselves show one or two possibilities at each critical point, namely, firstly, a very gradual transition from one phase to another, so that the observation points within about a 2 per cent, limit could equally well be represented on either of the lines ; secondly, a transition stage, in which the observation points lie upon neither of the straight lines, but one generally lower than both. Examples of each of these two possibilities will be discussed seriatim.
Firstly, a gradual transition. An example of this can be illustrated from the results obtained at 24'2 for acids of percentage values from 30-17 to 33*87, and in the following table (p. 98) the observed results for the contractions are compared with those calculated both from the 21 per cent, and the 33'3 per cent, origins of co-ordinates.
It is evident that the result for the 30-17 per cent, acid belongs to the straight line drawn from the 4 per cent, origin, and the result for 33-87 per cent, acid, on the other hand, belongs to the straight line drawn from the 33-3 per cent, acid, while those intermediate could within the degree of experimental error be equally represented as belonging to either line.
