Having eliminated cellulose nitrate as a possible fuel, Beller turned his attention to the refrigerant gas hypothesis. According to Robert Moulton's 1962 account of the fire, air-cooling units in the Melody
Lounge were served by a refrigerating unit behind the false wall, and
after the fire, some of its tubing had been found broken or melted. As
Moulton notes, however, none or the commonly used refrigerant gases are flammable, so this would seem rule our any refrigerant gas as being in any way responsible tor the initial flash."
Moulton also notes that, although some refrigerant gases used at the
time were toxic, there wasn't any reason to assume that the refrigerating
unit's tubing melted or broke to release the gas during the first
minutes of the tire. "It thus appears," he said, "that refrigerant gas may
be dismissed as a factor in the early stages of the fire when most fatalities occurred."
Or so it appeared in 1962. In 1993, however, David Arnold published
an article in The Boston Globe that seemed to shed doubt on this conclusion. According to Arnold, "Methyl chloride is a flammable gas that was commonly used as a refrigerant during the war years. It replaced freon, almost all of which was allocated to the military."
He goes on to say that, "it was common knowledge that the Coconut Grove was cooling beer, food...and people in the summer...with methyl chloride in system with a capacity of 10 to 15 tons." Apparently, investigators at the time thought the Cocoanut Grove was using freon or an older cooling chemical, sulfur dioxide. They didn't know about the methyl chloride, so it wasn't mentioned in their 1943 published report. No one noticed the omission because those who'd serviced the Coconut Grove cooling system never saw the report."