112 THE PERIODICAL CICADA. 
Almost at the moment that it becomes free, the larva begins to 
run actively alxmt with the quick motions of an ant, l)ut soon goes 
to the side of the limb, loosens its hold, and deliberately falls to the 
ground, its specific graAaty being so slight that it passes through the 
air as gently as a feather and receives no injurs'. The peculiar 
instinct which impels tliis newly hatched larva to thus precipitate 
itself into space without the least knowledge of the distance to the 
ground or the result of its venture has been often commented upon, 
but is not more remarkable than other features in the life liistoiy of 
tliis species. 
On coming to the earth the larva immediately penetrates it, usually 
entering at a crack or fissure, or at the base of some herbaceous plant, 
and begins the long period of its subterranean life. 
The newly hatched larva (fig. 46) is about one-sixteenth of an inch 
long and differs considerably in general form from the later larval 
stages, while at the same time presenting the general structural char- 
acteristics showm in the latter. It has, for example, a much longer 
and distinctly eiglit-jointed antenna, 
and the head is longer in proportion 
to the bod}^. It is 3^ellowish white in 
general color, except the eyes and the 
claws of the anterior legs, which are 
Fig. 46.— Newly hatched lan^a, greatly reddish. It is sparselv covered ^^dth 
enlarged. (Alter RUey.) miuute hairs. In f omi^ it is quitc elon- 
gate and subcylindrical, and it is particularly notable for its very 
prominent front legs. 
THE UNDERGROUND LIFE OF THE CICADA. 
EXPERIMEXTAL PROOFS OF THE LOXG UNDERGROUND LIFE. 
The life and habits of the periodical Cicada above ground, wliicli 
have so far only been discussed, are open to easy study and have been 
fairly well understood, certainly since the time of Hildreth, Potter, and 
Smith; but from the time of the disappearance of the young larva be- 
neath the soil and thereafter, throughout its long h;\^ogeal existence, 
observations are difficult and vdi\\ the earlier observers were limited to 
the occasional or accidental uneartliingof specimens, and no consecutive 
series of observations of a defuiite brood or generation was attempted. 
The proof for the 17-year or 13-year period for the development of 
the Cicada was, therefore, based solely on chronological records, but 
so noteworthy were the recurrences of the important broods and so 
fidl and complete were the records, some broods ha^ang been regularly 
recorded on the occasion of each visit for nearly two hundred ^^ears, 
that there was no possibility of doubting the accuracy of the time 
periods from such records alone; nevertheless, this unusual feature in 
