214 
BULLETIN OF THE BUREAU OF FISHERIES 
during life in which it lost its snout from the nostrils forward. The injury became completely 
healed over; and, notwithstanding that it had not even rudiments of jaws, the mouth being repre- 
sented by an oval-shaped opening, this fish not only survived but at the time of its capture was 
fatter and apparently in better condition than many normal individuals. 
It has been difficult to determine upon what tins fish feeds, for the stomachs examined con- 
tained mainly finely divided, flocculent substances, without definite shape or form. A few stomachs 
contained some cycloid scales, one contained fragments of shells of mollusks, and a few contained 
w r hat appeared to be strands of algae. It seems probable that the food is finely divided before it 
reaches the stomach. It is possible that grinding the food is a function of the teeth situated in 
the oesophagus. Butterfish taken at Woods Hole were found feeding on small fish, squid, crusta- 
ceans, annelids, and ctenophores. 
Spawning takes place chiefly during June and July in Chesapeake Bay. The earliest date 
when we observed ripe fish was May 26. Fish as small as 145 millimeters ( 5 % inches) had well 
developed roe late in May, indicating that maturity is reached at a length of about 6 inches. 
Apparently, the butterfish spawns throughout most of its range, for Bigelow and Welsh (1925, 
p. 247) give the season for the Gulf of Maine from June to August. The eggs are transparent, 
round, and 0.7 to 0.8 millimeter in diameter, and hatch in less than 48 hours at a temperature of 
65° F. 20 The larvae are 2 millimeters long at the time of hatching. At 15 millimeters the tail is 
forked, the dorsal and anal fins are formed, and the fry can be identified readily. 
While butterfish spawn in Chesapeake Bay, no very young fish were collected, due in part, 
at least, to the fact that tow r nets seldom were used during the summer. At the time when large 
catches of adults w r ere being made with pound nets none were taken near-by along shore with 
our collecting seines. It was noted throughout the collecting that butterfish rarely were taken 
within a few hundred feet of the shore. Large numbers of unmarketable fish, 3 to 5 inches long, 
taken the middle of May with pound nets, probably were about 1 year old. In the lower York 
River, on October 13, 13 butterfish 128 to 155 millimeters (5 to 6 inches) long were seined, and in 
the evening of that same day the catch was 10 of 135 to 165 millimeters (5.3 to 6.5 inches) in length, 
indicating that the growth from May to October is from about 4 to 5J4 inches. It is probable that 
by the following spring, when 2 years old, the fish will have reached a length of about 7 inches and 
maturity. 
20 For an account of the embryology and larval development of the butterfish see Kuntz and Radeliffe, 1918, pp. 112 to 110, 
figs. 58 to 68. 
