After another week of patient groping, Broffin was obliged to confess that the problem of identification was too difficult to be solved on conventional lines. It presented no point of attack. With neither a name nor a pictured face for reference, inquiry was crippled at the very outset. None of the many boarding- and rooming-houses he visited had lost a lodger answering the verbal description of the missing man. Very reluctantly, for bull-dog tenacity was the detective's ruling characteristic, he was forced to the conclusion that the only untried solution lay in Teller Johnson's unfortified impression that the chance meeting at his wicket was not the first meeting between the robber and the young woman with the draft to be cashed. It was the slenderest of threads, and Broffin realized sweatingly how difficult it might be to follow. Assuming that there had been a previous meeting or meetings, or rather the passing acquaintance which was all that the young woman's later betrayal of the man made conceivable, would the writer of the accusing letter be willing to add to her burden of responsibility by giving the true name and standing of the man whose real identity--if she knew it--she had been careful to conceal in the unsigned note to Mr. Galbraith? Broffin read the note again--"a deck-hand, whose name on the mate's book is John Wesley Gavitt," was the description she had given. It might, or it might not, be an equivocation; but the longer Broffin dwelt upon it the more he leaned toward the conclusion to which his theory and the few known facts pointed. The young woman knew the man in his proper person; she had been reluctant to betray him--that, he decided, was sufficiently proved by the lapse of time intervening between the date of her note and its postmark date; having finally decided to give him up, she had told only what was absolutely necessary, leaving him free to conceal his real name and identity if he would--and could. Having come thus far on the road to convincement, Broffin knew what he had to do and set about the doing of it methodically. A telegram to the clerk of the _Belle Julie_ served to place the steamer in the lower river; and boarding a night train he planned to reach Vicksburg in time to intercept the witnesses whose evidence would determine roughly how many hundreds or thousands of miles he could safely cut out of the zigzag journeyings to which the following up of the hypothetical clew would lead. For, cost what it might, he was determined to find the writer of the unsigned letter. XVIII THE ZWEIBUND