The truth is, that the offence and the punishment of Mattioli were well known to European diplomatists and readers of books. Casal, moreover, at this time was openly ceded to Louis XIV., and Mattioli could not have told the world more than it already knew. But, for some inscrutable reason, the secret which Dauger knew, or was suspected of knowing, became more and more a source of anxiety to Louvois and Louis. What can he have known? The charges against his master, Roux de Marsilly, had been publicly proclaimed. Twelve years had passed since the dealings of Arlington with Marsilly. Yet, Louvois became more and more nervous. In accordance with commands of his, on March 2, 1682, the two valets, who had hitherto occupied one chamber at Exiles as at Pignerol, were cut off from all communication with each other. Says Saint-Mars, ‘Since receiving your letter I have warded the pair as strictly and exactly as I did M. Fouquet and M. Lauzun, who cannot brag that he sent out or received any intelligence. Night and day two sentinels watch their tower; and my own windows command a view of the sentinels. Nobody speaks to my captives but myself, my lieutenant, their confessor, and the doctor, who lives eighteen miles away, and only sees them when I am present.’ Years went by; on January 1687 one of the two captives died; we really do not know which with absolute certainty. However, the intensified secrecy with which the survivor was now guarded seems more appropriate to Dauger; and M. Funck-Brentano and M. Lair have no doubt that it was La Riviere who expired. He was dropsical, that appears in the official correspondence, and the dead prisoner died of dropsy. As for the strange secrecy about Dauger, here is an example. Saint-Mars, in January 1687, was appointed to the fortress of the Isles Sainte-Marguerite, that sun themselves in the bay of Cannes. On January 20 he asks leave to go to see his little kingdom. He must leave Dauger, but HAS FORBIDDEN EVEN HIS LIEUTENANT TO SPEAK TO THAT PRISONER. This was an increase of precaution since 1682. He wishes to take the captive to the Isles, but how? A sedan chair covered over with oilcloth seems best. A litter might break down, litters often did, and some one might then see the passenger. Now M. Funck-Brentano says, to minimise the importance of Dauger, ‘he was shut up like so much luggage in a chair hermetically closed with oilcloth, carried by eight Piedmontese in relays of four.’