Theory the Second: Macdougall’s experimental protocols were as lame as his poetry. Let’s look a little closer at his findings.
Macdougall weighed six patients in all, but only the first, the one described earlier, stands as a strong example of the phenomenon. Macdougall threw out Patient 6’s data because the man died just as they had put him on the cot and were adjusting the beam. Number 4’s data he discounted because, Macdougall wrote in American Medicine, “our scales were not finely adjusted and there was a good deal of interference by people opposed to our work.” The doctor makes several references to “friction on the part of officials,” and states that only the first patient was run under ideal conditions, i.e., sans friction. He doesn’t specify what form this friction took, but if it affected the tests to the extent that some were thrown out, it seems reasonable to assume that the officials were there in the room, hectoring Macdougall or trying to bring a halt to what he was doing. Hardly ideal conditions for a test that requires concentration and enough quiet to listen for a heartbeat.
That leaves four subjects. With the exception of Number 1, the data for all were compromised in one way or another. Number 2 stopped breathing at 4:10 a.m., but the scale didn’t budge for another fifteen minutes (whereupon it registered a half-ounce drop). “We had great doubt, from the ordinary evidence,” writes Macdougall in his American Medicine paper, “to say just what moment he died.” If you can’t tell when the man died, you can’t very well claim that he lost a half ounce at the moment of death.
Number 3’s weight loss happened in two phases: a half-ounce loss at the moment of death, and then an additional loss of an entire ounce a few minutes later. Macdougall explains that the second loss might have been caused by a jarring of the scale, caused when one of his colleagues listened to the subject’s heart. If pressing a stethoscope to a patient’s chest disturbs the balance, as of course it would, then how did Macdougall and his colleagues presume to know the moment of death in any of these cases?
Number 5’s data were tainted by a peculiarity of the scale. Following a three-eighths-ounce loss, three-eighths ounces of weight was added to the scale to bring it back to zero; however, the beam didn’t budge for fifteen minutes. Macdougall had no explanation. Was his scale dodgy? Did Fairbanks make a reliable scale? Was it really accurate to one-fifth of an ounce? Where was a Fairbanks scale historian when you needed one?
PEGGY PEARL OVERSEES the Historical Collection of the Fairbanks Museum in St. Johnsbury, Vermont, where Franklin Fairbanks began manufacturing scales in 1830. The collection includes 30 or so antique Fairbanks scales, as well as farm implements, “tools of yesteryear,” 103 years of Northern New England Weather Center records, and the Carlton Felch diaries. Peggy fielded my call with a vigor that suggested things were pretty quiet around the Historical Collection office. The Fairbanks company’s scales, she said, were ubiquitous from 1830 through the first half of the twentieth century. They were the Rolls-Royce of platform scales. When I told her Macdougall used a silk scale with a capacity of 300 pounds, she faxed me two pages of Fairbanks Silk Scales from a Macdougall-era catalogue.
“It combines great sensitiveness with the increase in capacity and platform,” said the proud Fairbanks copy. “Very handsome in appearance.” Indeed, as Macdougall had said, his scale was accurate to one-fifth of an ounce. I told Peggy the story of Macdougall and the dying consumptives, hoping she might have some nugget of Fairbanks scale minutiae that would explain the good doctor’s three-quarter-ounce drop. She initially wondered whether the platform might have been inlaid, in which case, you couldn’t have a cot sticking out over the sides of the platform without rendering the results, as she put it, “screwy.” But the scale in the drawing had a standard suspension platform.
“That’s about as far as I can take you,” said Peggy. You could hear the disappointment in her voice. Peggy Pearl could tell me what the weather was on April 10, 1901, and she could tell me what Carlton Felch was up to that day, but she could not tell me whether Macdougall’s cot contraption had somehow compromised the Fairbanks’s renowned accuracy.
MACDOUGALL SEEMED AWARE of his study’s weaknesses, and he encouraged others to try to extend and replicate his work. He wanted to do more trials himself, but was stymied by the earlier-cited frictions with officials. Overtures to “positioned and entrenched scientific authorities” at other facilities, he writes in a letter to R. Hodgson of the American Society for Psychical Research (ASPR), were met by rebuffs. His best bet would have been the ASPR itself. Indeed, the ASPR’s scientific officer, Hereward Carrington, upon hearing of Macdougall’s study, wrote at length and with great gusto, in an issue of the society’s journal, about the possibility of outfitting a condemned prisoner in an airtight glass hood, and placing him, electric chair and all, onto a platform scale.
In the end, Macdougall resorted to weighing some dogs on a scale he set up in his barn. Owing to the difficulty of finding dogs dying from a disease that rendered them exhausted and motionless, he immobilized and then killed them via injection—fifteen dogs in all. Not one evinced a drop in weight as it died. Macdougall’s spin on this rather striking batch of conflicting data was that of the churchgoing Christian: Animals don’t have souls—or anyway, sayeth the Bible, not the eternal variety—and therefore we should expect this.
Not everyone in the soul-weighing business would agree with Macdougall there. Ten years after the American Medicine paper was published, a physics teacher at Los Angeles Polytechnic High School self-published a book called The Physical Theory of the Soul, which included a chapter detailing his adventures in mouse-soul-weighing. H. LaV. Twining, as he appears on the title page, didn’t seem to like animals much, as we will see, but he gave them credit for arriving in this world with the same spiritual accoutrements as humans. “It is reasonable to conclude…” he wrote, “that all forms of life have accompanying souls… and that animals would form fit subjects… since they could be killed at will and under any chosen conditions, while human beings could not.”
Over the following four pages, H. LaV. offs thirty mice on his scale, using every condition the Polytechnic supply closet would support. He suffocated them in test tubes melted shut by Bunsen burners (no weight loss). He smothered them in flasks sealed with rubber stoppers and flasks sealed with parafinned corks (no loss, and no loss). He gassed them in open flasks with cyanide pellets. Here at last he witnessed a loss: one to two milligrams “at its last kick.” H. theorized that the poison had caused the mice to “perspire violently at death,” and that the lost milligrams were evaporated mouse sweat. I did a little reading on the subject of cyanide poisoning, in the form of a paper by Dr. John M. Friedberg, which helped prompt the Ninth Circuit Court of Appeals to deem cyanide executions cruel and unusual. Death by cyanide does appear to be moderately aerobic—panic, retching, seizures, violent head extension, grimaces—though excessive perspiration isn’t specifically mentioned as a symptom. Excessive salivation is, however. Perhaps some seizure-flung drool escaped the confines of the flask.