The variation between March and June and September is irrational, and further examination suggests that it is false, depending on low figures returned for pigs <2 in those months. If the actual litter size on those dates was as in September (which, in that it was the basis for the rationing, is likely to have been exceptionally accurate) than the number of pigs <2 should have been, March, 914, June, 781, instead of 796, and 708 respectively. Pigs under 2 months on any quarter are 4 or 5 months of age next quarter, and, apart from deaths and changes of rate of production, represent 2/3 of the total pigs between 2 and 5 months on that day.