134 
however, a fish is attacked, the whole of the lower series of teeth is 
everted, thrown forward out of the mouth, and each set of two rows 
widely separated on the mid-line; they then present an apparatus 
much like a grappling-devil used in dredging, and are illustrated on 
Plate XIII, fig. 2. 
Describing the American Polistotrema stouti, Jordan and Ever- 
mann write:* “The hagfish fastens itself usually on the gills or 
isthmus (throat) of large fishes, sometimes on the eyes, whence it 
works its way very rapidly into the inside of the body. It then de- 
vours all the flesh of the body without breaking the skin, so that the 
fish is left a living hulk of head, skin, and bones. It is especially 
destructive to fishes taken in gill-nets. In every gill-net in summer, 
at Monterey, more or less of these empty shells of fishes are obtained. 
When these are taken from the water the hagfish scrambles out with 
great alacrity. It is thought that the hags enter the fishes after they 
are caught. A fish of 10 lb. to 15 1b. weight will be devoured by 
them in a single night. Large fishes of even 30 lb. weight are often 
brought up without flesh and without viscera, and they certainly 
do not swim into a gill-net in this condition.” 
When trawling in the neighbourhood of Otago Heads, and par- 
ticularly in Blueskin Bay (which, from the many diseased fishes ob- 
tained there, is dubbed “ the Hospital” by local fishermen), quite a 
number of fishes were obtained, chiefly terakihi (Cheilodactylus macro- 
pterus) and red-cod (Physiculus bachus), with large wounds in their 
sides: these wounds, I am convinced, were the work of blind-eels ; 
and, as the fishes were still alive, there can be no doubt that these 
parasites were feeding when enclosed by the net, or attacked the 
fishes already enclosed, and, being disturbed when hauling the trawl, 
left their victims and escaped through the meshes. In no instance 
did I observe that the gills had been attacked, though it is quite pos- 
sible that ingress may be also obtained in that way. What seems 
worthy of emphasis is the circumstance that the blind-eel can actually 
hold on to the side of a fish, rasp off the scales, and bore into the flesh. 
Whether this is possible only when the victim is netted, or whether 
a free fish can be so attacked, we have not sufficient evidence to show. 
It is well known that fishes caught in a meshing or gill net are par- 
ticularly liable to attack: one such instance was afforded at the 
Chatham Islands. A meshing-net set for moki (Latris ciliaris) on 
being raised was seen to have a large tangled knot in it. On clearing 
the knot it was found to contain a terakihi and a blind- eel, both dead ; 
the former with a large wound in its side, the latter almost inextri- 
cably threaded through the meshes, and in one place almost cut in 
two by the twine. The one surpassing circumstance was the fact 
that the blind-eel was meshed at all, considering its eel-lke shape 
and slimy coat. 
* Jordan and Evermann, Bull. U.S. Nat. Mus., 47, 1, 1896, p. 6. 
