[Marston's seniors] were not reassured by his performance in his first and only real-world case. When surgical instruments and a military codebook vanished from the U.S. Surgeon General's office in Washington, D.C., intelligence officials suspected one of the building's eighteen [not seventy, as Marston later claimed!] African-American messengers. Asked to find the culprit, Marston identified one man "with very strong consciousness with regard to something he had done," but he also admitted that he was not certain of the man's guilt. The problem, he explained, was that blacks, perhaps because their will was more "primitive" than that of whites, seemed to respond differently: "The factor of voluntary control which, with white men, seems to make a deception rise regular and almost an absolute one, apparently is almost altogether lacking in negroes, so that, tho the change is really even more sharp and extreme, it is vastly more difficult to estimate norm plus excitement."

...Mr. Bruce Bielaski, then Chief of the Bureau of Investigation, Department of Justice, arranged for me to work with the Bureau's New York office which had many puzzling spy problems. One of the most interesting cases concerned the loss of a secret code book. The code had disappeared from the office of the Surgeon General, in Washington, and would be taken, investigators believed, to German agents in New York City. There were upwards of seventy negro messengers in the Mills Building, Washington, which housed the Surgeon General's offices. Any one of them might have gained access to the room where the code key had been hidden. There was nothing for it but to test all seventy negroes, eliminating from suspicion those whose test records were clear.

Here was a job indeed, a lie detecting task which looked impossible on the face of it because of the extreme emotionality of negroes. A very slight impulse will send a colored man's blood pressure sky-rocketing upward, while sudden fear may cause it to drop just as suddenly again to the depths. Nevertheless, I decided to test all the messengers. Nearly all of them were desperately afraid of the test, believing it to be some sort of magic or voodoo. Some of the messengers had to be brought into the examining room by force. When all had been tested and their records examined I did find a few who hadn't stolen anything. But the number of petty thefts which our deception test uncovered was amazing--every imaginable sort of loot from pencils to fur coats and surgical instruments were among the articles recovered.

Despite all these complications, however, it was surprisingly easy to spot the messenger who had taken the code key. I telephoned immediately to the official in charge of the investigation. But although he thought all entrances to the building were covered by government agents the guilty messenger managed to get away through the basement. This was just as well, as things turned out, because the thief was trailed to the German agent who had hired him in New York City and the code book was recovered. This type of elimination test with the Lie Detector has since been used repeatedly in police investigations--it solves the otherwise impossible problem of narrowing suspicion among large groups of people.

Following the successful solution of several spy cases with the Lie Detector, Major Robert M. Yerkes, one of America's leading psychologists and Chief of the Psychological Division of the Sanitary Corps, suggested that I might organize a corps of deception testers in his division. These specially trained operatives might be assigned later to spy cases and other deception problems in Army Intelligence, Naval Intelligence, and the Bureau of Criminal Investigation, Department of Justice. I agreed to undertake the work and accepted a second lieutenant's commission to carry it out. In the course of our preliminary applications of the deception test a good many military cases were successfully handled. But while I was still at Camp Greenleaf training a group of army psychologists in the use of the test, the Armistice was declared. This put an end to war-time development of the Lie Detector. It left the Marston Deception Test approved by science as a reliable and practical method of detecting guilt or innocence, but not as yet accepted by law enforcement officers or by the Courts.

Major R.M. Yerkes, of the Surgeon General's Staff, recommended the blood-pressure test for practical use, with breathing and galvanometer records to be made and synchronized with the blood-pressure record for the sakes of supplementary check on the lie-catching expert's judgments....Extreme skepticism met Major Yerkes' endorsement of the Lie Detector in Army circles. Then it chanced that the code book was stolen from the Surgeon General's office in Washington. Circumstances made it evident that one of the seventy negro messengers in the Mills Building had probably taken the all-important code key, probably at the instigation of a German spy. I was called upon to examine all seventy negroes with the Lie Detector. Some of those colored boys had to be herded into the examining office at the point of a gun! They were terribly afraid of the test, every one of them. They thought it was black magic, or some voodoo procedure. As the test proceeded it became apparent that a good many of them had practical reasons for being afraid---we uncovered an amazing list of articles which had been stolen from various offices. Here was an excellent opportunity to test out another criticism of the Lie Detector--namely its alleged failure when suspects or witnesses were afraid or emotionally upset from causes having nothing to do with the problem under investigation.

The Lie Detector stood the test. None of the negroes, terrified as they were, beat the lie-catching expert. The messenger who had stolen the code book was found, and the book traced and recovered. More important from a long-run point of view was the fact that another alleged method of beating the test had been disposed of. You cannot beat the Lie Detector by letting fear run riot in your brain.