29. The next step in the analysis is to treat separately the seats held by the two parties respectively. In Table XVII this distinction is observed. The number of seats, I beg leave to observe, is entered to indicate the position of the group-majority in relation to the general party aggregate, and not as connoting any relation between it and the majority of which it is the result. As this table is based upon my general compilation, the figures for unopposed seats, adjusted as I have already described, are included. The effect of the adjustment upon the total will be found discounted at the foot of the table.

30. To analyse this table in full detail would be beyond the scope of the present paper. I will comment therefore on the main features only. Taking the county constituencies first, it will be observed that the Unionists, with a gain of 38 seats, increased also their majority. The increase runs through all the groups to a greater or less degree except where one seat only was gained, namely, in the East and Northern Midlands. On the other hand the Gladstonians seem to have lost not only the 38 seats, but, to some extent, their majorities even in the constituencies which were not diverted from their allegiance. The falling off is very marked in three of the four groups north of Trent, a fact which is all the more striking alongside of the great increase in the majorities by which the seats of this party were held or won in the Lancashire and West Midland groups. The number and relative importance of the unopposed returns in the latter must be taken into consideration. In Lancashire we find Clitheroe and Rossendale; in Staffordshire, Burton; in Gloucestershire, the Forest of Dean. There are also the two divisions of Monmouth, where the majority, at least in one of the two, assumed quite Welsh dimensions. The general average is raised by these items, so that the divergence from the average of 1892 is comparatively small, although it is in marked contrast to the corresponding figure on the other side. The adjustment of uncontested seats affects, necessarily, the Unionist majority more than the other, as the latter show only 6 to 74 in possession of the former. The case is reversed in 1892, when these adjustments were of less importance to either side. Finally we have the difference in the majorities between the seats retained and those lost. Here again whilst the Unionist majority advanced from the already sufficient proportion of 40 per cent. to 66, the Gladstonian proportion, which was still more ample in 1892, fell from 47 to 29. The figures regarding the transferred seats again present interesting features. Put in one way, three seats held in 1892 by the Unionists with a majority of only 18 per cent., fell to the opposite party in 1895 by a majority of 42. Looked at from the

point of view of votes given to each separately, the Unionist poll in these seats increased by 3 per cent., whilst the Gladstonian waxed and multiplied by over 9. When we turn to the seats lost by the latter, which were 41 in number, it appears that the majorities were on a more liberal scale in both elections. In 1892 the seats were won by a majority of 12 per cent., which was transformed on the last occasion to one of nearly the same amount on the other side In other words, the party in possession at the election of 1895 lost about 7 per cent. of its votes of 1892, and the assailant's success was due to an increase in its poll amounting to 17 per cent.

31. In the case of the boroughs the tendency was much the same. The nine seats won in 1895 by the Gladstonians by an average majority of 58, were held by the opposite party in 1892 by a majority of 75. So again the Gladstonians held in 1892 52 seats by a percentage majority of about 18, which succumbed to the Unionists on the later occasion, by a majority averaging just under 12 per cent. In the case of the first mentioned transfers, the Unionists showed a poll, deficient as compared with that of 1892 by a little under one per cent., whilst the Gladstonian poll grew by nearly 13. The difference in the other case is more marked, for the poll of the party in possession fell off by over 7 per cent., and that of the Unionists gained by 22. In the rest of the borough return there is a good deal more irregularity than in the county portion of the table. The general tendency of the Unionist poll was strongly upward, and that of the others slightly downward. The large Unionist majorities in the metropolis were increased, but the 8 Gladstonian seats show a falling off in this respect, amounting to one-half the proportion of 1892. In the provinces, we may compare the curious inconsistency of a gain of seats by the Gladstonians in the South West group of boroughs, with a diminished majority, and in the extreme North, an equally striking instance of the same kind on the Unionist side, indicating the closeness of the struggle, especially when we note that the other seats held by the losing party in the same groups rallied with unusual vigour to their banner. Look, for instance, at the North, where the Gladstonian poll in the 7 seats left to it, ran to a majority high above that of 1892, whilst at the other end of the country the Unionists raised their majority in their 6 seats to 16 per cent., whereas, when they held 8 seats, it was only 11. In the South Midlands, too, the Gladstonians kept by 44 per cent. I seat, as against 4 by 38 per cent. in 1892. There is an apparent anomaly, to which I may direct attention, in the case of the borough group of the Eastern counties, where the Gladstonians exchanged an unopposed seat in Norwich for a hard won victory

at Ipswich, whilst Yarmouth and Colchester, so far as majorities are in question, only balanced each other proportionally, though not in the actual number of votes. The difference between the voting in London and the provincial boroughs makes it advisable to show the latter separately in respect to the seats held and transferred, a point which I could not conveniently take into consideration in framing the table on which I have just been commenting. In the following statement this omission is supplied:

TABLE XVIII.-Boroughs, including London.

32. Before leaving the subject of boroughs, I should like to revert for a moment to the general division of votes, irrespective of the party by whom the seat was held, in order to indicate how far the different classes of boroughs were in harmony with the territorial and general totals. Table XIX gives the main facts in the usual proportional form.

The poll shows a general increase, except in the first group, where there is a fractional falling off owing to the small opposition offered to the two Gladstonian seats in Hull and Bristol respectively, which were not contested by Unionists. Here, however, as in the rest of the groups, the Gladstonian poll alone diminished, though that of the opposite party rose by less than elsewhere. The net results of the voting are then given. Only one group showed a Gladstonian majority in 1892, and this gave place at the next election to one of a trifle more pronounced character on the other side. I am inclined to attribute this to the fact that in 1892, out of eight unopposed returns among these two-member constituencies, six were those of Gladstonians, whereas in 1895 the number, although unchanged, was equally divided between the parties. The insignificant Unionist majority in the middle-class borough constituencies rose proportionally more than the rest, though the majorities in both the largest and the smallest groups were actually larger. On the whole, excluding London, the majority on

this side was about six times the proportion of that of 1892. The small towns, though showingno change in the total results so far as representation is in question, had four changes. Two of the scats passed from Unionist to Gladstonian, and two executed a movement in the opposite direction. On the total of all groups, outside the metropolis, 10 seats in every 25 held by the Gladstonians in 1892, passed over to the enemy in 1895. The Unionists increased their representation by 27 per cent., and though their majority grew relatively more than this, it was not remarkably high.

33. The uncontested seats must now make their appearance for, I hope, the last time, at all events as a prominent feature of the returns I am discussing. I have already shown, first that they are no new incident in a general election, and again, that, though since the reconstruction of parties in 1886 they have been more numerous on the Unionist side, in the earlier days of the reforms of 1867 they varied with the wave of popular support. On the present occasion we have to deal with only 13 on one side and 27 on the other in 1892, but in the succeeding contest, the Unionists increased their number to 110, and the other party were left with one less than their former total. Of the 169 county seats held by the Unionists in the latter election, 44 per cent. would have had to be omitted from my returns had some adjustment not been made. Of the 42 borough seats uncontested, only 6 are Gladstonian, and of these, 4 represent the second seats in the two-member boroughs, and the others are those of Sheffield, where no contest took place for any one of the 5 seats. Of the Unionist borough seats of this class, 11 were in London, and the rest scattered, small and large towns alike contributing. Birmingham and Liverpool on the one hand, Taunton, Windsor, Winchester, and Bury St. Edmunds, on the other. I may mention that of the 122, 26 were left uncontested in 1892, as well as on the later occasion. Excluding these from TABLE XX.-Uncontested Seats of 1895.

consideration, over 45 seats were held at the former election by majorities of more than 50 per cent., and only 16 by less than