The first Boston series ended on 23 December 1918 having produced nothing. Six weeks later the same people went back and did it again, with the design altered to answer the objections their own report had raised.
The men and the place
On 4 and 6 February 1919, forty-nine volunteers from the Naval Detention Training Camp at Deer Island crossed to the quarantine station on Gallops Island. The newspapers said fifty, and said they had been chosen from several hundred who offered. The Associated Press called them a death squad; the Washington Herald reported that “every man at Deer Island volunteered to enter what they termed the ‘death squad.’”
Six of the forty-nine never came under experiment. That left 43 men, who took 82 inoculations between them: one man had three, thirty-seven had two, and five had one.
They were older than the first contingent: two were nineteen, thirty were between twenty and twenty-five, nine were between twenty-six and thirty, and two were thirty-three and thirty-six. They weighed between 125 and 182 pounds. The mean weight on admission was 157 pounds and on discharge 157.6, which tells you something about how the island was run.
The report describes Gallops Island carefully this time: sixteen acres, six miles down Boston harbour, a quarter of a mile from the nearest land at Fort Standish on Lovells Island, about thirty buildings with modern heating, lighting and sanitation. The winter was unusually mild. During the experiments the temperature ran between 18 and 50 degrees Fahrenheit, “as a rule the days were clear, and plenty of sunshine prevailed,” and there was always a brisk breeze that sometimes rose to a forty-mile gale. “No one, at any time, experienced any inconvenience, much less hardship, from the weather during the sojourn on this station.”
For the first week the men were quartered together in barracks, ate at the same mess, “were allowed to congregate at will,” played outdoor games and did light chores. For five days before anything was done to them, their temperatures were taken at half past eight in the morning and half past six at night.
That observation week is the most useful thing in the second series, and it was not an experiment at all. Between 5 and 10 February, before a single inoculation, twelve of the forty-nine men reported sick with tonsillitis of varying degrees. Three were admitted to hospital with sore throat, headache and malaise. One more, number 18, became ill the day he arrived and ran an intermittent fever for twelve days that three doctors examined on 13 February and could not name; they decided it could not be diagnosed as influenza “though the possibility of an atypical attack could not be entirely ruled out,” and he was taken off the experiment.
So before the second series began, a quarter of the volunteers had sore throats, and one had a fever nobody could explain. This is the population in which the experimenters proposed to detect influenza by clinical judgement.
What was changed
Two things.
The material would be fresher. The first series had used secretions collected hours earlier, and its discussion had admitted this might be fatal to the result: “In no case, however, did we obtain material during the period of incubation.” The second series went as early as it could get. In Experiment VII the material was taken in the ninth hour after the donor’s first symptoms and instilled within fifteen minutes of leaving him. In Experiment VIII it had been recovered within four hours of a man’s first symptom.
They added passage experiments. If a volunteer seemed to have caught something, his secretions were taken and given to the next group. A disease that can be passed from a man you gave it to, on to a third man, is a living agent; an illness that stops with the first man is a coincidence. Experiments IV, VII and IX were passage experiments.
They also went after a second candidate organism. In 1918 a Chicago bacteriologist, George Mathers, had reported a streptococcus from influenza cases at Camp Meade which some thought was the true cause. Four of Mathers’s strains were obtained through Dr Hektoen and pooled, and Experiment I, on 11 February, sprayed 1.5 cubic centimetres of them into the nasopharynx of ten men. Three of the ten had their noses made alkaline with sodium bicarbonate first, and three had them made acid with acetic acid, in case the chemistry of the nose was what had been defeating them.
“The results were entirely negative during a period of seven days observation.” One incidental finding from that experiment is worth recording: the proportion of volunteers carrying Pfeiffer’s bacillus in the nose and throat rose from 40 per cent before the inoculation to 80 per cent seven days after it, which is a fair demonstration that the bacillus spreads easily between men who are living together and has nothing to do with whether they are ill.
The two cases
Across nine experiments, two men fell ill in a way that could not be dismissed.
Number 5, aged twenty-three. In Experiment II, on 13 February, ten men had three cubic centimetres of nasopharyngeal washings put into the nose, taken two hours earlier from a doctor who had fallen ill the previous evening — the donor is described only as Dr A. C., who had been treating influenza patients all winter without catching it, and who left a banquet at half past eight on 12 February with a pain in his back and a temperature of 100.8 degrees.
Five days later, number 5 lost his appetite, went to bed after lunch, and by evening had a headache, chills over the whole body, backache, weakness and a worse cough. His temperature reached 39.2 degrees on the third day. He had a flushed face and a mottled pharynx and no tonsillar involvement at all, which mattered, because tonsillitis was what everyone else was getting. The report’s heading for his case reads: “Diagnosis.—Influenza (?). Incubation period.—Five days (?).”
It adds: “The case so closely simulated one of influenza that a passage experiment (No. IV) was done, using this patient’s secretions as the source of the material.” That passage experiment produced tonsillitis in one man and nothing else.
The report’s own weighing of the case is a model of its kind: “In favor of influenza is the presence of headache, backache, depression, and exacerbated cough, while against it is a relatively high leucocyte count.”
Number 42, aged twenty-one. In Experiment VII, on 24 February, ten men were inoculated with nasopharyngeal secretions from a supposed early case. Number 42 is described as a man who had always enjoyed good health but “had a nervous disposition and was inclined to be timid and introspective,” who had had typhoid at ten and pneumonia at fifteen, who had never had influenza and had only casual contact with influenza patients.
He said afterwards that he had not felt well since the inoculation. He reported sick two days later with no fever. On the fourth day his temperature was 37.3 degrees, the next evening 37.8, and he was admitted with a headache and a cough that had come on suddenly. The following day his temperature went to 39.2 in the morning and 39.9 in the evening. He complained of a splitting headache, backache, pains in the chest, abdomen and limbs, and a horror of light. His face was flushed, his conjunctivae injected, his throat a dusky red with no exudate, his chest clear.
On 3 March Lieutenant Commander McGuire of the Navy Medical Corps examined him, found crackles over both lower lobes, and said — in the sentence the whole programme comes down to — that “in an epidemic the case would surely be called influenza.”
The report records the incubation period as “sixty-six to one hundred and twenty hours,” which is a way of saying that nobody could pin down when it started.
What the second series actually transmitted
Streptococcal tonsillitis, reliably, repeatedly, and in a form that could be proved by culture. The charts drawn for the report by the laboratory’s artist, Leonard H. Wilder, show it plainly: pooled material in a circle, lines running out to numbered boxes, and under them the results, which read “One case of tonsillitis,” “Tonsillitis after 40 hours,” “Tonsillitis 36 hours after,” “Tonsillitis 2 days after with secondary febrile attack.”
The conclusions say so, and this is the most useful finding of the whole programme:
In contrast to the difficulty in transmitting influenza by means of secretions, acute streptococcic tonsillitis may readily be transmitted in this way, and with a high percentage of success, even when the donor is apparently merely a carrier of the streptococcus.
The conclusions
Four of them, published in 1921:
That the results “indicate presumptively that influenza may be transmitted by means of the secretions of the upper respiratory passages from patients in the early stages of this disease, probably within less than 12 hours from onset.”
That nothing definite could be concluded, for two reasons: “the uncertainty of our diagnosis in recipients and donors on account of the lack of decisive criteria as to what is influenza,” and “the clouding of our results by the transmission of streptococcic tonsillitis to many of our volunteers.”
That what success there was came “in only a small percentage of the instances attempted, the recipients being young male adults in a region where epidemic influenza had recently prevailed, and possibly, therefore, of more than average resistance.”
And that “attempts to transmit influenza by means of cultures of Pfeiffer’s bacillus and of Mather’s streptococcus were unsuccessful.”
That last line is the one that outlived the rest. The Boston experiments never gave anybody influenza, but between them they put Pfeiffer’s bacillus up the noses of more healthy men, in larger doses, from more strains, than any other experiment of the period, and nothing whatever happened. For a bacillus that had been the accepted cause of influenza for twenty-six years, that is not nothing. It is the one firmly positive thing a programme of negative results produced, and it is negative too.