With the store-combs the six angles of the cell fulfil an equally important office. The ideal honey-cell would be one with its mouth opening upwards, so that it could be filled in an ordinary rational way. But under the strict economical principles ruling in the hive such an arrangement would be impracticable. The honey-vats must be stacked one over the other in a horizontal position, and therefore must be chargeable from the end. All cells in the comb have a slight upward tilt, but not enough to retain the fluid contents if the cell were a round one. The effect of the angles in the hexagon is to increase the retentive property of the cell, and experience has taught the bees how to supplement this natural holding power of the angles by just that slight cant of the cell which is necessary to prevent the nectar running out. The worker-bee, during her period of larval life, at first lies coiled up at the bottom of the cell, but as her size increases she takes up a position lengthways, with her head towards the cell-mouth. This, however, is not a constant attitude, for she seems at intervals to make a series of slow gyrations or somersaults, probably to facilitate the casting of her skin, which she accomplishes several times during her five days’ life as a grub. At the end of this time the nurse-bees stop the feeding process and seal up the cell. Now the larva sets to work, first to spin herself a silken shroud before entering on her long sleep as a chrysalis, and then to change her skin for the last time. In the case of the worker these fine-wrought sleeping-clothes envelop her whole body, forming a continuous cocoon. But the queen-larva weaves herself only a scanty sort of cloak, covering her head and thorax, but leaving her nether portions bare. The theory usually advanced in explanation of this is, that when the surplus queens are slaughtered in their cells by the accepted mother-bee after her fertilisation, the fell work is rendered easier by the absence of the tough material of the cocoon over the parts generally attacked. It seems to be well substantiated that in a battle of queens the stings are not used haphazard, as with the workers, but each queen tries to thrust her weapon into one of her enemy’s spiracles or breathing-holes, of which she possesses fourteen, seven on each side. And a stroke dealt in this way appears to be always fatal.