in September) he had already begun to solicit the king's permission to retire from office, pleading ‘the comfortless prospect of very ill-health for the future.’ On this occasion he was prevailed upon by the king and his intermediary, Portland, to remain (Correspondence, pp. 6–14). In December he showed his fidelity to the whigs by seeking to dissuade the king from proroguing the Convention parliament with a view to its dissolution. When, early in 1690, it had been dissolved and succeeded by a parliament where the tories preponderated, and showed themselves indisposed to accept, unless in a hopelessly mutilated form, the abjuration bill warmly advocated by him, his resolution to resign became fixed (Burnet, iv. 81). In spite of the king's repeated refusals to accept his resignation and Tillotson's remonstrances, Shrewsbury sent back the seals by Portland on 3 June 1690, after having been dissuaded with difficulty by Burnet from making his way into the royal presence in order to speak his mind (ib.; cf. Correspondence, pp. 16–17 and note; Correspondence, &c. of Henry, Earl of Clarendon, 1828, ii. 316; Memoirs of Queen Mary, ed. Doebner, 1886). The answer to the question whether ‘temper’ or orders from St. Germains determined Shrewsbury's resignation depends on the general opinion to be formed of his conduct during the ensuing four years.
From June 1690 to March 1694 Shrewsbury remained out of office, maintaining a general attitude of opposition to the measures of the tory ministers. On the arrival, however, of the news of the disaster of Beachy Head (30 June 1690), he hastened from his retirement at Epsom to offer his services to Queen Mary, proposing to raise troops (Dalrymple, iii. 87, 99), and declaring his readiness to take the command of the fleet, should it be assigned to some great nobleman, with two experienced seamen to advise him (Shrewsbury to Caermarthen, ib. pp. 130–1; cf. Macaulay). In January 1693 he was one of the eleven peers who protested against the renewal of the act for subjecting literary publications to the control of a licenser. About the same time he came forward as the mover of the triennial bill, to which, although almost unanimously favoured by the lords, the opposition of the tories in the commons encouraged the king to refuse his assent (ib. chap. xix.). But a few months later misfortunes both by sea and land determined the king to throw himself once more upon the whigs; and on his return to England in November he took the seals of secretary of state from Nottingham, and personally offered them to Shrewsbury. The interview, however, ended unsatisfactorily, and Shrewsbury withdrew to Eyton, his seat in Oxfordshire. An effort to induce him to change his mind was now made by Elizabeth Villiers, the king's mistress, with the aid of a daughter of Robert Lundy [q. v.], the former governor of Londonderry, to whom Shrewsbury was attached. But, though their endeavours were seconded by some of the whig leaders, it was not until some months later, and after other whig appointments had been made, that Shrewsbury (4 March 1694) again accepted the secretaryship of state (Correspondence, pp. 19–30).
His return to office has, however, like his previous resignation, been thought to have had a hidden reason. According to Macaulay (chap. xix.) both these actions on his part were due to the change which had come over him with the dissolution of the Convention parliament, when his allegiance to the new régime had first begun to waver. He now, it is said, entered into relations of the most compromising character with the court of St. Germains; and it was by the direction of James II that in 1690 he resigned his secretaryship of state. So it was stated in a memorial submitted by James to Louis XIV in November 1692, and included in the ‘Nairne Papers,’ afterwards published in Macpherson's ‘Original Papers’ (i. 435). Elsewhere in the same series of papers his name stands forth conspicuously in the so-called ‘Melfort Instructions,’ which were conveyed by or through his mother, the Countess of Shrewsbury, to himself, Marlborough, and Russell [see Drummond, John, titular Duke of Melfort]. The chief purpose of these ‘instructions’ was to secure to Russell the command of the fleet, while Shrewsbury was to help to retard its sailing as long as possible (Original Papers, i. 456–7). His name was again prominent in a paper supposed to date from the last quarter of 1693, and giving a list of King James's chief supporters at home (ib. p. 459); and in Lloyd's account, stated to have been delivered at Versailles on 1 May 1694, this agent professed to have been assured by the Countess of Shrewsbury that her son had returned to office only when he had been informed by King William that he had cognisance of Shrewsbury's discourses concerning King James, and after having retired into the country with the design of joining the latter should he land in England; this expectation had broken down. But though he had thus again taken office
