At first, all went according to rule, the mother digging and the father clearing away the rubbish. A few droppings were stored; then, on reaching the bottom of the test-tube, the couple pined away and died. The layer of sand was not deep enough. Before piling the food-sausage on top of an egg, the pair needed a shaft at least forty inches in depth, whereas they had only some eighteen inches to dig in. This failure did not put an end to my list of questions. Where did that pigmy spring from? Was he the outcome of a special predisposition, transmitted by heredity? Or was he descended from another dwarf, who himself proceeded from a similar abortion? Was his deficiency merely an accident, which had nothing to do with heredity, an individual littleness not transmissible from father to son? I incline to the theory of an accident. But what sort of accident? I can think of only one liable to diminish the size without injuring the type: I mean, a lack of sufficient food. We argue thus: animals virtually take shape in a mould whose capacity may be extended in proportion to the amount of molten substance which the crucible pours into it. If this mould receives only the strictly necessary amount, the result is a dwarf. Anything beneath this minimum means death by starvation; anything above it, in doses which increase but are soon limited, means a prosperous life and a normal or slightly larger size. The bulk is decided by plus or minus quantities of food. If logic be not a vain delusion, it is therefore possible to obtain dwarfs at will. All that we need do is to diminish the provisions to the lowest limits compatible with the maintenance of life. On the other hand, we cannot hope to make giants by increasing the ration, for a moment comes when the stomach refuses any excess of food. Natural necessities may be likened to a series of rungs of which the one at the top cannot be passed, while it is quite practicable to stand higher or lower on those near the bottom. First of all we must discover the regular ration. The majority of insects have none. The larva grows up amidst an indeterminate supply of victuals; it eats as it pleases and as much as it pleases, with no other check than its appetite. Others, those most richly endowed in maternal qualities such as the Dung-beetles and the Bees and Wasps, prepare definite rations of preserved food, neither too large nor too small. The Bee stores up in receptacles of clay, cement, resin, cotton or leaf-cuttings just the right amount of honey for a larva’s welfare; and, as she knows the sex of the future insects, she puts a little more at the service of the grubs that are to become females and will be slightly larger and a little less at the service of the grubs that are to become males and therefore will be smaller. In like manner, the Hunting Wasps dole out their game according to the sex of the nurslings.