On the other hand, the entomologist may desire to try the effect of a continued _low_ temperature on his pupae. This he can do by placing his pupae in an ice house. Such experiments have often been performed, and the results are very interesting. In some cases the emergence of the perfect insect has been delayed for many months, and even years; and then, after an exposure to a normal temperature lasting only a week or two, the winged insect has made its appearance just as if nothing unusual had happened. Such are the effects of _extreme_ temperatures on the duration of the chrysalis state; and we naturally infer, from such results, that the pupa under natural conditions is influenced, though in a lesser degree, by the variations experienced with the seasons, especially in such a fickle climate as our own. The insect hunter has always to bear this in mind, and particularly so when he sets out on a search for certain desired species. Suppose, for example, he has set his mind on the capture of a certain butterfly that _usually_ appears in the _first_ week in May. Before finally naming the day, he has to consider what the weather has been during the last few weeks, and if he finds that this has been much warmer than the average for the corresponding periods in the past, he selects a day in _April_, earlier or later according to the difference between the present season and the average. If he does not pay due attention to such considerations, he will sometimes find that all the insects netted are shabby and much worn, even though, under average conditions, he would be catching newly emerged and brilliant specimens. This, then, will explain how it is that we so often see in entomologists' periodicals startling accounts of 'early captures,' and of the appearance of certain insects late in the season that _ought_ not to have emerged till the following summer. I will give just one illustration of these variations. The beautiful Orange Tip Butterfly (Plate I, fig. 7) generally appears about the middle of May in the southern counties. Farther north it is of course a little later. In the north of England it has been taken in June; and in Scotland as late as July. On the other hand, I have taken it in Gloucestershire as early as March, on a rather bleak day with a cold east wind; and, withal, in a field on the slope of the Cotswolds fully exposed to the unfavourable breeze; but it was evident that, in this case, the butterfly had been enticed from its winter quarters by the milder weather of the few previous weeks. It may be as well, in passing, to observe that it is not only the pupa that is influenced by temperature. The hatching of eggs may be forced by high temperatures, or be retarded by exposure to cold; and in nature the period of incubation varies with the seasons. The larvae, too, grow faster or more slowly, or pupate earlier or later from similar causes. And so no very definite date or period can be assigned to any one stage of any insect.