metamorphosis." We have seen that even in them the term "complete" is a relative and not absolute expression, and that the terms larva and pupa are convenient designations for states varying in duration, and assumed to fulfil certain ends of existence, and even then dependent on length of seasons, variation in climate, and even on the locality. When we descend to the insects with an "incomplete" metamorphosis, as in the May fly, we find that, as in the case of Chloeon, Sir John Lubbock has described twenty-one stages of existence, and let him who can say where the larval ends and the pupal or imaginal stages begin. So in a stronger sense with the grasshopper and cockroach. The adult state in these insects is attained after a number of moults of the skin, during each of which the insect gradually draws nearer to the final winged form. But even the so-called pupae, or half winged individuals known not to be adult, in some cases feel the sexual impulse, while a number of species in each of the families represented by these two insects never acquire wings. [Illustration: 205. Larva. 206. Semi-pupa. 207. Advanced Semi-pupa. 208. Pupa. EARLY STAGES OF THE HUMBLE BEE.] Still how did the perfect metamorphosis arise? We can only answer this indirectly by pointing to the Panorpa and Caddis flies, with their nearly perfect metamorphosis, though more nearly allied otherwise to those Neuroptera with an incomplete metamorphosis, as the lace-winged fly, than the insects of any other suborder. If, among a group of insects such as the Neuroptera, we find different families with all grades of perfection in metamorphosis, it is possible that larger and higher groups may exist in which these modes of metamorphosis may be fixed and characteristic of each. Had we more space for the exposition of many known facts, the sceptic might perceive that by observing how arbitrary and dependent on the habits of the insects are the metamorphoses of some groups, the fixed modes of other and more general groups may be seen to be probably due to biological causes, or in other words have been acquired through changes of habits or of the temperature of the seasons and of climates. Many facts crowd upon us, which might serve as illustrations and proofs of the position we have taken. For instance, though we have in tropics rainy and dry seasons when, in the latter, insects remain quiescent in the chrysalis state as in the temperate and frigid zones, yet did not the change from the earlier ages of the globe, when the temperature of the earth was nearly the same the world over, to the times of the present distribution of heat and cold in zones, possibly have its influence on the metamorphoses of insects and other animals? It is a fact that the remains of those insects with a complete metamorphosis (the bees, butterflies and moths, flies and beetles) abound most in the later deposits, while those with an incomplete metamorphosis are fewer in number and the earliest to appear.