Mr. Knapp describes a still more curious structure of another species of ant common in this country:--"One year," says he, "on the third of March, my labourer being employed in cutting up ant-hills, or tumps as we call them, exposed to view multitudes of the yellow species (_Formica flava_) in their winter's retirement. They were collected in numbers in little cells and compartments, communicating with others by means of narrow passages. In many of the cells they had deposited their larvæ, which they were surrounding and attending, but not brooding over or covering. Being disturbed by our rude operations, they removed them from our sight to more hidden compartments. The larvæ were small. Some of these ant-hills contained multitudes of the young of the wood-louse (_Oniscus armadillo_), inhabiting with perfect familiarity the same compartments as the ants, crawling about with great activity with them, and perfectly domesticated with each other. They were small and white; but the constant vibration of their antennæ, and the alacrity of their motions, manifested a healthy vigour. The ants were in a torpid state; but on being removed into a temperate room, they assumed much of their summer's animation. How these creatures are supported during the winter season it is difficult to comprehend; as in no one instance could we perceive any store or provision made for the supply of their wants. The minute size of the larvæ manifested that they had been recently deposited; and consequently that their parents had not remained during winter in a dormant state, and thus free from the calls of hunger. The preceding month of February, and part of January, had been remarkably severe; the frost had penetrated deep into the earth, and long held it frozen; the ants were in many cases not more than four inches beneath the surface, and must have been enclosed in a mass of frozen soil for a long period; yet they, their young, and the onisci, were perfectly uninjured by it: affording another proof of the fallacy of the commonly received opinion, that cold is _universally_ destructive to insect life."[CO] The earth employed by mason-ants is usually moist clay, either dug from the interior parts of their city or moistened by rain. The mining-ants and the ash-coloured (_Formica fusca_) employ earth which is probably not selected with so much care, for it forms a much coarser mortar than what we see used in the structure of the yellow ants (_F. flava_) and the brown ants (_F. brunnea_). We have never observed them bringing their building materials of this kind from a distance, like the mason-bees and like the wood or hill ant (_F. rufa_); but they take care, before they fix upon a locality, that it shall produce them all that they require. We are indebted to Huber the younger for the most complete account which has hitherto been given of these operations, of which details we shall make free use.