28 The American Naturalist. [January, 
seen, in part, even without destroying the shell, since, when the : 
crab is not too active, the anterior part of the Polydora may pro- 
trude from the external opening in the columella, reach about in — 
the water, and presumably find and swallow food in the sand or — 
currents of water. Yet it is very easily disturbed, and then to be 
found only within the columella, or even in the upper calcareous — 
tube, where it may be variously coiled about, since the dwelling A 
is everywhere wide enough for the Annelid to bend back and 
forth upon itself. The part of the old shell utilized by the An- 
nelid is thus intermediate between that the crab occupies and — 
that covered over by the Hydroid. The former, in drawing in — 
and out of the shell, tends to limit the excursions of the Annelid — 
by dragging its claws over the columella and orifice of the 3 
Annelid’s tunnel ; the latter extends up to the very edge of the 4 
Annelid’s place of exit. Sometimes the Annelid’s tunnel appears — 
to have been cut through into the chamber occupied by the her- 
mit, but then covered in again by a calcareous layer protecting: : 
the Annelid from contact with the crab. 
Considering how impossible it would be for the Annelid to — 
keep the shell free from sand and prevent its burial, it is obvious — 
that this Polydora, if it lives, as it appears to, only in such a com- — 
mensal state, is dependent upon the habit of the hermit crab, and _ 
thus exists in what is a somewhat recently acquired environment. — 
As, however, other species of Polydora are known to make tun- 
nels into various Gasteropod and Lamellibranch shells, dead or © 
alive, it is evident that this particular species has not departed — 
so widely from the habit of its kind. 
Before describing the interesting breeding habits of this com- 
mensal Annelid, we will give an account of its structure and of © 
certain organs especially illustrated in the figures. 
The body (Fig. 2) is about a millimeter wide, about twenty-five ~ 
long, and rather flat, with little color except the bright red of the 
blood in the conspicuous dorsal vessel and in the branchia. 3 
These reach half way across the back, and each contains a vascu- 
lar loop that passes from the dorsal vessel to the tip of the 
branchia, then back again into the body. The limb of this loop — 
nearer the middle line or to the dorsal surface may possibly act ; 
