Four times, between 21 February and 9 April 1896, Ministers were asked – thrice by Rollit – what steps were to be taken in the Session to implement the Scheme of reconstruction. Each time, either Gorst or Balfour simply replied that the matter was under the consideration of the Lord President: on 9 April, Gorst told Rollit that the Duke was abroad and there would be ‘no final decision until he returns’. But as early as 10 February, Wynne had been ‘led to understand that the Duke would like to know how the country members voted’, on 21 January. Wynne asked Gregory Foster, at UCL, to let him have an analysis. ‘Two sets of numbers should be prepared, one regarding all those outside the administrative County of London as country members, and a second regarding all those resident beyond say twenty-five or thirty miles as country members.’

Foster’s analysis does not seem to have survived, but, several months later, Bennett sent Wynne his version of the vote, in which he had included ‘outlying districts such as Kew Richmond Croydon Blackheath and Woolwich as parts of London’. The analysis showed that there was doubt whether twenty-six voters were Londoners or countrymen. But of the others present on 21 January, 384 Londoners and fifty-eight countrymen were in favour of reconstruction, and 181 Londoners and fifty countrymen were opposed. No doubt there were many analyses, and certainly there was continuous correspondence in the press, both sides being well represented. In Manchester, the opponents set up a Manchester District University Defence Association. The actual numbers were often mere accompaniments to a conviction well expressed by one opponent of reconstruction, who claimed, with almost certainly a high degree of accuracy, that the vote did not ‘represent the views of the majority of . . . members [of Convocation], but rather the views of those engaged in teaching in the metropolis’.

While it is unlikely that the Duke of Devonshire was greatly interested in other aspects of the vote, it is worth looking briefly at two breakdowns by degrees.