----"each crawling insect holds a rank Important in the plan of Him who framed This scale of beings,--holds a rank, which, lost, Would break the chain, and leave behind a gap Which Nature's self would rue." The duration of the insect in the pupa state, though variable, is, without doubt, limited, and sometimes it is fixed to _an hour_, quite irrespective of all external circumstances. The most remarkable example of this kind occurs in the case of the insects whose larva we have already mentioned--the _Ephemera_. These insects appear with the greatest regularity, issuing from the waters of the Seine or Marne, in France, between the 10th and 15th of August. The fishermen call them _manna_; and when their season is come, they say, "the _manna_ begins to appear," or "the manna fell abundantly last night," alluding, by this expression, either to the astonishing quantity of food which the insects afford to the fish, or to the large quantity of fish which they then take. The fishermen expect them with the greatest confidence during these few days, nor are they ever disappointed. Millions upon millions suddenly rise into the air between eight and ten o'clock in the evening, and this generally for three successive nights.[N] Whatever be the temperature of the atmosphere, whether it be cold or hot, these flies invariably appear at the same hour in the evening, that is, between a quarter and half-past eight; towards nine they begin to fill the air; in the following half hour they are in the greatest numbers; and at ten there are scarcely any to be seen. So that in less than two hours--and these always the same--this infinite host of insects leave their pupa state, become perfect insects, perform their appointed work, and vanish. The same phenomenon of regularity of limit to the pupa state occurs also in other insects, though, perhaps, less strikingly. Some insects constantly leave the pupa at break of day; others in the full tide of noon-day, and others when the shadows of declining day come over the landscape. These, however, are certainly exceptions to the general rule, which appears to lay down no precise period of the day or month when this state is ended, and the perfect state is entered upon; but a limit, nevertheless, exists, mainly dependent for its appointment upon the external influences of warmth and air.