One of the most striking features of the Northern Ireland referendum was the scale of the turnout. The numbers voting last Friday exceeded by 220,000 or 30 per cent the turnout at last year's General Election.

Who were these additional voters? Will they turn out again in the Assembly election four weeks hence? And, if so, for what parties might they vote?

The RTE exit poll provides important clues to the answers to these questions.

This analysis is based on the intentions of the 95.6 per cent who disclosed their voting intentions. It is probable that there could be a very small understatement of the Sinn Fein vote as its supporters are less inclined to reveal their allegiance to pollsters.

The table below sets out the voting pattern in last year's general election and the voting intentions of the much greater number of people who voted in this referendum.

In interpreting these figures it should be borne in mind that in an election some parties put up candidates only in certain constituencies, whereas in a referendum their supporters in every constituency vote. This must account for some part of the increased support recorded in this table for the Ulster Democratic Party (which fielded no candidates in the general election), the UK Unionists (who put up only one candidate, Bob McCartney) - and the Progressive Unionists, Women's Coalition and Workers' Party, all of which last year fielded candidates in only a small number of constituencies.

The first thing that is clear from this table is that in terms of numbers voting, Protestants and Catholics shared almost equally in the increased turnout - with around 110,000 more of each tradition voting than last year.

However, this represented a much bigger increase in the Catholic than in the Protestant vote - almost two-fifths as against just over a quarter.

Next, at first sight there may not appear to have been much difference in the turnout by both communities, for the nationalist parties' share of the vote, at 40.9 per cent, is close to the 4142 per cent which is the best approximation of the Catholic share of the adult population in Northern Ireland.

But that may be somewhat misleading because, for the reasons given earlier, the Sinn Fein vote is probably slightly understated. Moreover, earlier polls suggest that about one-fifth of Alliance voters and perhaps one-third of those voting for the Women's Coalition and other small cross-community groups are Catholic.

When allowance is made for these factors, it emerges that it is possible that as many as 45 per cent of those who voted were Catholics.

And if that were to have been the case, then the Catholic turnout would have been almost 90 per cent and the Protestant turnout only 75 per cent.

Now, this may well exaggerate the difference in turnout between the two communities, but on the basis of the available data it is very difficult to resist the conclusion that there was such a differential. And I have to say that something of the kind always seemed likely in view of the extraordinarily high proportion of the unionist community, and especially of UUP voters, who were shown by the polls to have been undecided right up to the time of the referendum.

The Irish Times poll of May 20th showed one-sixth of all unionist voters and almost a quarter of UUP voters still undecided - as against 1 per cent of nationalists. Against that background it was clearly unlikely that the turnout of unionists would match that of nationalists.

A LOWER turnout of UUP voters would help to explain the extraordinary drop in UUP support which seems to be suggested by the exit poll. The proportion of those voting in the referendum who declared themselves UUP supporters as they left the polling stations was barely half the proportion who voted UUP in last year's general election. As a result, despite the overall high poll, the actual number of referendum voters who stated an intention to give their first preference to a UUP candidate in the forthcoming Assembly election was 68,000 less than in the 1997 general election.

It must be added, however, that a low turnout of UUP supporters could not explain the whole of this drop in the UUP share of the vote. It is impossible to resist the conclusion that a substantial proportion of UUP voters have turned away from the party, at least temporarily.

Where have they gone? Some may have defected to Bob McCartney's UK Unionists, whose support grew markedly during the campaign. But quite large numbers would seem to have moved their allegiance to cross-community groups - Alliance but also the Women's Coalition and other small parties. For this group of parties, which are neither nationalist nor unionist, appear to have collected more than twice as many votes as in last year's general election - a scale of increase that could not be accounted for wholly by the failure of the smaller groups to present candidates in many constituencies last year.

The picture facing the UUP is, therefore, by no means as bleak as these data might at first sight seem to suggest. First of all, most of the party's supporters who stayed at home in the referendum - and that number may have been substantial - will probably vote for it in the Assembly election.

Second, Protestant votes for small cross-community parties will in most cases fail to elect a member even in a relatively large six-seat constituency, and many of these votes will then pass on to the UUP. And, third, some at least of those who told pollsters on referendum day that they would vote "Other" may well have been UUP voters expressing what may have been temporary disgruntlement with that party, rather than intending to vote for any specific small party.

Finally, it is clear from these figures that the No parties were successful in motivating to vote in the referendum a large number of people who do not normally take part in elections: the total number of referendum voters supporting the DUP or UK Unionists was more than 50 per cent greater than in the general election. And, given the scale of the Yes victory, many of these may not be motivated to turn out again to vote for these parties in the Assembly election.

A final point worth making is that within the extreme unionist family the UK Unionists have certainly gained ground against the DUP. Paisley's appeal seems to be fading as he loses support on the one hand to the small loyalist parties and on the other hand to McCartney's UK Unionists. The fact that among those who voted in the referendum less than 13 per cent intended to vote DUP must be worrying for its ageing leader.