280 
BULLETIN OF THE BUREAU OF FISHERIES 
smelts were later introduced from Swan Lake, where a large “race” or “size” exists. 
The latter were planted in Mooselucmaguntic. 
Again, there is Square Lake, in Aroostoko County, where no small “race” or 
“size” has manifested itself as adult breeding fish, but smelts up to 10 inches or 
more are common. These smelts were derived from Swan Lake stock, according to 
L. T. Carleton, formerly commissioner of inland fisheries and game. 
Still again, Sunapee Lake, in New Hampshire, affords an example of waters 
stocked with smelts that seldom, if ever, attain a large size. At least there are no- 
breeding runs of large smelts. The source of the smelt eggs planted in Sunapee 
Lake is not definitely known, but probably they were either from Squam or Winne- 
pesaukee Lake. The majority of the smelts now taken in Sunapee Lake are small, 
running about the size given for smelts of Winnepesaukee Lake; but even these little 
smelts are caught on baited hook in deep water and have been known, as elsewhere 
stated, to do likewise after spawning while still remaining in the brook. Nothing of 
this sort is known of the small smelt of Sebago Lake, which appear to subsist exclu- 
sively upon plankton; so if these fish originated in Squam Lake they have degener- 
ated in size. That such is the case is indicated by a note in Forest and Stream of 
October 23, 1890, under the title of “Sunapee Lake Fishing,” which says: 
The smelt which was introduced more than 20 years ago occasionally reaches 10 inches in 
length. It is extremely abundant, and is often washed up on the eastern shore of Sunapee by 
strong winds. 
On the other hand, Cheney (1896c, p. 429) wrote that in November, 1896, 
while walking on the beach of Sunapee Lake, N. H., with Commissoner Wentworth,, 
they picked up a number of smelts thrown up on the sands by the high winds and 
water. These were of two sizes, one from 1% to 2 inches long and the other 3 
to 3 inches long. They regarded the smaller size to be of the previous spring’s- 
hatching, while the larger ones were supposed to be adults that, had they lived, 
would have spawned the following year. 
Still further evidence in this direction is that only the “large” form of Green Lake 
smelt was introduced into Crystal Lake, Mich., for the reason it was the large form 
only that was propagated at that station at that time. Creaser (1925) indicates 
that there was a spawning run of large fish only at Crystal Lake. The smallest 
smelts secured, which were 2-year-old fish, were 6)^ inches in total length. Older 
fish had attained a length of 12 inches. 
The forgoing are rather insufficient data upon which to generalize, but they 
suggest that in some waters there are “races” that do not attain a large size under 
any circumstances, and that the large race sometimes decreases in size so as to 
appear like the small race; but in such instances the habits are like those of the 
large race. 1 
HABITS OF FRESH-WATER SMELTS 
FOOD AND FEEDING 
It has been stated that the food of the fresh-water smelt varies according to the 
size of the fish, and it may be added, according to its age. The character and 
arrangement of its teeth indicate its carnivorous propensities, while its comparatively 
