The ovipositor frequently attains a great length in these Insects, so as to exceed that of the body. It is used in different ways, some of the family depositing their eggs in the earth, perhaps in vegetable matter under the surface; but other species place the ova in twigs or stems of plants, arranging them in a very neat and compact manner in two series, as depicted by Riley[245] in the case of _Microcentrum retinerve_ (Fig. 190). These eggs are laid in the autumn, and in the following spring become more swollen before hatching. The Insect undergoes a moult during the process of emerging from the egg. By the time the emergence is completed the _Microcentrum_ has expanded so much {314}in size that it is a matter of astonishment how it can ever have been packed in the egg; the young commence jumping and eating leaves in a few minutes. Including the ecdysis made on leaving the egg, they cast their skins five times. The post-embryonic development occupies a period of about ten weeks. The larvae eat their cast skins. When the final moult occurs the tegmina and wings are at first quite soft and colourless, but within an hour they assume their green colour. These Insects, as remarked by Riley, make interesting pets. The people of the Amazon valley are in the habit of keeping a species in cages, and our British _Locusta viridissima_ does very well in confinement. One of the most curious habits of these Locustidae is a constant licking of the front paws. Riley says that _M. retinerve_ bestows as much attention on its long graceful antennæ as many a maiden does upon her abundant tresses, the antennæ being drawn between the jaws and smoothed by the palpi. This American naturalist also tells us that he reared three successive broods in confinement, and that the Insects gradually deteriorated, so that the eggs of the third generation failed to hatch. [Illustration: FIG. 190.—Eggs of Katydid (_Microcentrum retinerve_): A, the two series at deposition; B, side view of a single series. (After Riley.)] The ovipositor, which is one of the most characteristic features of the Locustidae, is not present in the newly-hatched Locustid (Fig. 191, A), the organ being then represented only by two papillae placed on the penultimate segment. The structure and development of the ovipositor in _Locusta viridissima_ have been described by Dewitz.[246] Fig. 191, A, shows the young Insect taken from the egg just as it is about to emerge. The abdomen consists of ten segments, the terminal one bearing at its extremity two processes, the cerci, _a′_. These persist throughout the life of the Insect, and take no part in the formation of the ovipositor. The tenth segment subsequently divides into two (_a_, _a'′_, Fig. 191, C), giving rise to the appearance of eleven abdominal segments, and of the ovipositor springing from the antepenultimate.