The probable error on one pig calculated on the average daily live weight increase was found to be for cross-bred pigs about 7 per cent.; for pure-bred pigs from the same litter and each of similar live weight about 3 per cent.; for cross-bred pigs each of similar live weight but from different parents about 4 per cent.
The error is greatest in experiments of short duration and it decreases as the period of the experiment is lengthened. The decrease is slight if the experiment is prolonged beyond seven weeks.
The errors on 18 separate pens containing four pigs to each pen (the total initial live weight of each pen being similar) varied from 2·3 per cent, to 12·08 per cent., the average being 6·21 per cent.
The percentage probable error is relatively low when the pigs are fattening rapidly and high when the pigs are fattening slowly. The average daily increase of the quicker fattening female pigs when un-speyed appears to be depressed as a result of periodic sexual disturbances.
The stage of fattening does not appear to influence the probable error. Animals in the later stages of fattening show a greater range in their rate of fattening than animals in the early stages, see Fig. 2, p. 284.
Weekly weighing of individual pigs is recommended as it gives an indication of the health of each pig. It also enables each experimenter to discriminate between differences in the rate of fattening due to individuality and differences due to ill health, etc.
The rate of fattening of pigs is very variable, periods of rapid fattening are followed by periods of depression.