THE SMELTS 
299 
the brook each successive day until May 1 . But there was plenty of time in which 
they could have migrated unobserved. 
It frequently has been stated and generally supposed that after spawing smelts 
invariably return to the lake on the night of their ascent. The writer’s observations 
on the marine smelt in small coastwise brooks revealed that, when undisturbed dur- 
ing the night, large numbers, if not all, remained in the brook the next day, and often 
some smelts lingered in the brooks long after the spawning season was over, becom- 
ing emaciated and weak. Those remaining after the spawning season, so far as 
examined, always proved to be males. 
These facts led to the suspicion that possibly fresh-water smelts might have a 
similar habit; and at Sunapee Lake it was found to be a fact that if the smelts were 
undisturbed during the night before the next day large numbers were found along 
Pike Brook as far up as they could ascend, but mostly congregated in the deeper pools. 
On April 16, 1910, notwithstanding the fact that there was some “dipping” during 
the first of the night before at the mouth of the brook, schools of smelts were found 
all along the brook, from just below the hatchery up 200 or 300 yards, in every little 
pool, and the same conditions obtained on the 17th. On the 20th smelts were 
observed in the pools, but there were not as many as could have been expected from 
the run of the night before. After the 20th no large numbers were observed during 
the day, but groups of a few or individuals here and there were sometimes seen (Ken- 
dall, 1914). 
At Sunapee Lake it was observed that smelts, sometimes at least, begin to feed 
before descending to the lake. On April 20, 1910, in a large deep pool, some smelts 
appeared to be feeding, moving leisurely here and there as though picking up or look- 
ing for something floating in the water. In the afternoon the writer, using a tiny 
hook with a small piece of earthworm for bait, caught six of the smelts, which proved 
to be spent or partly spent males, still having rather large milts. Two were 4, one 
4M> two 4}4> and one 4^ inches in length. There were many more bites, but the 
fish could not be hooked. Some of the fish would come up to the bait slowly, open 
their mouths, and take it in; some would dart at it quite smartly; some would not 
notice it unless it were moving rapidly; and some would pay no attention to it what- 
ever. The latter were the larger smelts. The stomachs of three of the larger fish 
caught contained smelt eggs and several insect larvae, apparently mosquito or possibly 
midge. 
SEX PREDOMINANCE 
The preceding notes suggest the predominance of male fish, at least during the 
early part of the season. According to Atkins, of 185 smelts taken April 26, 1875, 
131 were males and 54 females; but on the other hand, on April 21, 1869, of two 
dozen smelts caught the majority were females. 
Sixty-four smelts taken in a brook entering Panther Pond on April 17, 1910, 
comprised 48 males and 16 females. Of 526 smelts caught with hook and line in 
Sebago Lake, 218 were males and 308 females, but these were not breeding fish. 
The best available data on sex predominance are afforded by observations made 
at Sunapee Lake in 1910. The following table shows that in this instance males 
predominated during nearly the entire breeding season. 
