646 THE AMERICAN NATURALIST.  [Vor. XXXIX. 
close the cylinder upon its dorsal surface (Fig. 4). The circu- 
lar muscle by contracting around the long, forces it back to 
where it has more room in the hind end of the * club," and so 
helps it to withdraw the tooth plate. Considering the extreme 
cumbersomeness of this device for using the lower jaw, it is 
rather remarkable that the hagfish can move it with the light- 
ning-like rapidity with which it does. 
If the fish is large enough to permit of it, the hagfish makes 
only one opening in the skin, and pushing in through that, 
. works its way around inside, eating as it goes. I have seen 
three hagfish attack one fish through the same opening, their 
heads entirely hidden in the fish's body, their tails flapping like 
streamers in a wind as they pushed the fish in front of them, 
each striving to outdo its neighbor. They usually eat together, 
and I have often seen several of them at work on one dead fish, 
while other fish would be lying untouched in the aquarium. 
Under these circumstances it is no longer permissible to hold 
that the large fish found to be reduced to skin and bones, have 
been thus denuded by a single hag, though it may well be that 
only one of the number is detected in the act of leaving it. 
The hagfish do not eat often. After the first feeding in cap- 
tivity, a week passed before they were given anything more. 
Then flounders were given again, but only half a dozen ate. A 
week later, when flounders were offered them for the third time, 
only three touched them. At the end of another week they 
were given a rock cod, but did not touch it. Four days after, 
they were given a small cod and two small flounders, but did 
not touch them. It was not that the food did not suit them, 
for they eat any fish on the lines; they were evidently not 
hungry. StillI thought I would see if a different kind of food 
would tempt them more, so the next week, thirty-five days after 
they were caught, and seventeen days since any of them had 
touched food, and about a month after most of them had fed, I 
gave them five sardines, very flat fish, about eight inches long 
and two inches broad. There were about thirty hags in the 
tank, and by this time most all of them seemed to be hungry, 
for they ate all but one of the sardines. This was the only 
ume ravenous hunger was observed during the summer. The 
