332 BULLETIN OF THE BrKEAU OF FISHERIES 



Habits and food. — Although this fish is so easily recognized that fishermen 

 have long been familiar with it and scientists acquainted with it, little is known of 

 its mode of life. Certainly it is quite as voracious as its relatives, for it takes any 

 bait and is said to eat whatever invertebrates it finds on the bottom — e. g., moUusks 

 (both bivalve and univalve) , various crustaceans, sea urchins, and worms. Sea ravens 

 also eat fish, Vinal Edwards having found herring, launce, sculpins, tautog, silver 

 hake, and both sculpin and sea-raven eggs in specimens taken at Woods Hole. 



The sea raven alone among Gulf of Maine sculpins has the power of inflating its 

 belly like a balloon when lifted from the water. If thrown back again in this 

 condition it floats helplessly on its back, feebly waving its tail to and fro, and we 

 can not say whether it can deflate again at \vill like a puffer (p. 29S) or must await 

 the gradual escape of the air it has swallowed. Another point in which the raven 

 differs from our other sculpins is that it can bite sharply, due to its larger teeth. 



Off the southern shores of New England sea ravens work inshore in autumn and 

 out again into slightly deeper water in spring, but no seasonal movement of this sort 

 has been reported in the cooler waters of the Gulf of Maine; and apart from such 

 bathic migrations (which, after all, mean merely that shoal water is too warm for 

 their comfort in summer) sea ravens are resident throughout the j'ear wherever 

 found. 



Breeding haiits. — All that is definite^ known of its breeding habits (to which 

 we can contribute nothing first hand) is that ripe females have been foimd off' 

 southern New England in November and December; that the eggs are very large 

 (about 4 mm. in diameter), yellow when first spawned, soon changing to amber 

 color, and that they sink and stick together in masses.^' The period of incubation 

 is unknown. We can give no account of its larval stages, for the identity of the 

 cottid larvse referred to the sea raven by Agassiz and Whitman (1885) was not certain, 

 but in summer, when the young have grown to about 1^ inches (45 mm.), they are 

 to be found on the bottom. 



Presumably the sea raven breeds throughout its geographic range, but so far as 

 we can learn the Bay of Fundy is the only part of the Gulf of Maine where fry as 

 small as this have been definitely recorded. There, however. Huntsman found 

 them on both the New Brunswick and on the Nova Scotian shore in summer. 



Commercial importance. — Although the sea raven is said to be a good table fish 

 (we have never tried it) there is no more market for it than for other sculpins in 

 New England or Canada, but it is generally considered the best of all baits for 

 lobster pots, hence shore fishermen save what ravens they catch for this purpose. 



ALLIGATORFISHES. FAMILY AGONIDJE 



These curious little fishes are connected with tlie sculpins anatomically, though 

 their general appearance gives no hint of the fact. Their most striking external 

 feature is that the body is armed -with several rows of overlapping plates. The only 

 Gulf of Maine species somewhat suggests a pipefish in this and in its slender form, 

 but the mouth being of ordinary form there is no danger of confusing it with the 

 latter. Some agonids have a spiny dorsal fin ^nd others lack it, while the ven- 

 trals of all are thoracic. 



" Described by Bean (1903, p. 547). The pelapc eggs previously referred to this species (.\gassiz and Whitman, 1885, 

 p. 10) belonged to some other fish. 



