38 studies, every one cited on the page where it is used. Temperature is not something you add to a compost heap. It is the readout that tells you whether the four things you actually control are set right. Almost all composting advice inverts that. It gives you a temperature to hit, as though heat were an input — when heat is what a heap produces if, and only if, you have fed it correctly. That inversion is why “my heap won’t heat up” is the most common question in the subject, and why the usual answer (“turn it more”) so often fails. This is a plain-language, fully cited field manual built on that one idea, and it is honest about what the evidence does and does not support. What’s inside The rule everybody repeats, and what survived it. 55 °C for three days is in every guide and written into regulation. Salmonella, E. coli O157:H7 and Listeria have been recovered from poultry-manure compost held above 55 °C for more than eight days. Salmonella Typhimurium survived at least nine days at 60–70 °C. Why it fails — the measured mechanism, which inverts the usual advice. Pathogens warmed for one hour at 47.5 °C first, then held at 55 °C, lost 1.2, 1.9 and 2.3 log against 4.0, 5.6 and 4.8 in unwarmed controls. Roughly a thousandfold difference, at exactly the temperature the regulation specifies. Every heap has a warm-up, and the longest warm-ups belong to the worst-built heaps — so a heap that struggles to heat spends that time hardening the very organisms the heat is meant to kill. Where 30:1 came from. A 1953 experiment by McGauhey and Gotaas — and it is an optimum for speed. The ratio that loses least nitrogen is a different number (25:1), and every ratio from 20:1 to 30:1 matured at the same 80 days anyway. Plus why the published C:N figures for real materials are far too loose to calculate with. Particle size, the dial that pulls against itself. Grinding roughly doubles the decomposition rate and simultaneously closes the pores that carry the air. There is no setting that maximises both, which is why “shred everything finely” is wrong advice for a garden heap. Moisture, the failure nobody warns you about (a working heap dries itself out), and the fact that the optimum is a property of the material rather than a universal number. Why a small heap never heats — exact geometry, not opinion — including the uncomfortable arithmetic on the plastic bin most people are actually using. What turning really costs. Faster, and it loses moisture and ammonia. Both directions are measured; the exchange rate between them has never been published, so nobody can honestly give you an optimal interval. Weed seed. Barnyardgrass and pigweed die in three days at 49 °C. Field bindweed needed seven days at 82 °C, which no garden heap will ever reach. What compost actually is. Roughly 1.5–3.5% nitrogen against a bag labelled 10-10-10. An amendment with a nutrient side effect, not a fertiliser with a structural side effect — and that single fact explains a great deal of disappointment. Nitrogen drawdown — real, measurable at up to 17 kg N per tonne of wood, temporary, and overwhelmingly about placement rather than material. Compost tea and activators. The majority of the literature supports greater disease suppression by non-aerated tea, with aeration secondary to feedstock quality — so the part being sold is the part with the least evidence behind it. And the controlled trial closest to a garden heap found the uninoculated control did just as well as three different bio-inoculants. Worms, bokashi and leachate. Above 35 °C the earthworms die, so a worm bin has no thermophilic phase and cannot sanitise. Bokashi is fermentation, not composting. And the liquid in the tray is a fertiliser and an effluent, at around fifty times the strength of untreated sewage. It closes with A fault table arranged by what you can see, not by what is wrong A one-page reference to all four settings and the edges of each Every number in the book, collected and labelled where it is convention rather than measurement A glossary, and 38 numbered sources And a page listing what the research could not establish — because a book about unchecked numbers owes you its own gaps Format 49 pages, 12 figures. PDF (US Letter, print-ready) and EPUB (reflowable, validates with zero errors). Written for an international audience: metric first, with any country-specific regulation named as such. No fluff, no filler, and no invented numbers.