The mechanism of their spray works thus: Secretory cells produce hydroquinones and hydrogen peroxide (and perhaps other chemicals, depending on the species), which collect in a reservoir. The reservoir opens through a muscle-controlled valve onto a thick-walled reaction chamber. This chamber is lined with cells that secrete catalases and peroxidases. When the contents of the reservoir are forced into the reaction chamber, the catalases and peroxidases rapidly break down the hydrogen peroxide and catalyze the oxidation of the hydroquinones into p-quinones. These reactions release free oxygen and generate enough heat to bring the mixture to the boiling point and vaporize about a fifth of it. Under pressure of the released gases, the valve is forced closed, and the chemicals are expelled explosively through openings at the tip of the abdomen. [Aneshansley & Eisner, 1969; Aneshansley et al, 1983; Eisner et al, 1989]
Much creationist literature gives an inaccurate account of the process. Based on an admittedly sloppy translation of a 1961 article by Schildknecht and Holoubek [Kofahl, 1981], Duane Gish claimed that hydrogen peroxide and hydroquinones would explode spontaneously if mixed without a chemical inhibitor, and that the beetle starts with a mix of all three and adds an anti-inhibitor when he wants the explosion. [Weber, 1981] In fact, the two do not explode when mixed, as others have demonstrated. [Dawkins, 1987, p. 86-87] (Schildknecht did propose a physical inhibitor which kept the mixture from degrading in undisected beetles; in fact, the degradation he saw was probably simply a result of exposure to the air.) Gish still used the mistaken scenario after being corrected by Kofahl in 1978. [Weber, 1981] The same mistake is also repeated in books by Hitching in 1981, Huse in 1983 and 1993, and twice in a creationist magazine in 1990 [Anon, 1990a,b].