Chapter 10 of The Analysis of Burned Human Remains contains some interesting information on pyre cremations.
First, this chart of pyre temperatures:
Second, some notes on the influence of weather on pyre cremations:
The length of time required for the pyre to burn depends largely on the quantity of wood used to build it (the author has generally used about 700–900 kg; Holck, 1989: 43; McKinley, 1994a: 78–79) and, to some extent – particularly in temperate climates at certain times of year – on the weather (McKinley, 2006). Strong winds would result in the pyre burning faster and more fiercely but not necessarily more efficiently since burning would be uneven and result in the collapse of the structure (Figure 10.3); windbreaks can assist in cutting down the risk and attendants on hand could help redress the problem. Heavy rain would inevitably result in the cessation of cremation, whilst moderate and/or persistent rain would at the very least reduce the temperature.
There are indications that incomplete oxidation of the bone occurred more frequently and to a greater extent in the towns compared with the rural areas, overall both appearing less effective than the cremations undertaken in the Northern Frontier zones. In the towns, cremations would have been undertaken by professional ustores and payment would have been made for the quantity of wood to be used (Toynbee, 1971: 45; Noy, 2005). Inevitably, the poor, unable to afford sufficient fuel, appear to have been less well cremated than the better-off (Morris, 1992: 43), a problem still experienced in contemporary cultures (Barber, 1990: 380).
Roman cremations were probably mostly undertaken within two to seven days of death (Toynbee, 1971: 45; Noy, 2005). In Britain, particularly at certain times of year, this must have presented difficulties with regard to suitable weather conditions (McKinley, 2006). Fuel could be well seasoned and kept dry up until the point the pyre was lit, but a sudden unexpected downpour or the need to catch a break in the weather in prolonged periods of rainfall would, undoubtedly, have lead to problems at times, particularly in towns where the holding of a corpse for any extensive length of time would have potentially caused health risks as well as concern on the part of the relatives/associates of the deceased. These things did not always go according to plan and these pragmatic decisions that sometimes had to be taken are indicated by those less well-cremated remains recovered from several cemeteries, particularly, but not exclusively, those in the towns.