THE SMELTS 
333 
dried, they were eaten without cooking, or sometimes roasted on coals or in the 
oven. The fish dried were those taken in the spawning time. The writer once 
dried some taken in July, but they proved to be too oily and the flavor was 
unpleasant. 
In Maine the fresh-water smelt, besides being eaten, is often used as bait for 
landlocked salmon and other fish; and, as previously indicated, it is regarded as 
necessary to stock with smelt any body of water in which landlocked salmon are to 
be introduced, as the smelt forms its principal subsistence under natural conditions. 
FISH-CULTURAL PROPAGATION 
The earliest attempts to hatch smelts "artificially” in this country seems to be 
that of Charles G. Atkins in 1868, an incomplete account of which was given in his 
notebook, which the present writer had the privilege of consulting. Under date of 
May 5, 1868, he wrote: 
At the hatchery house I was greatly surprised to find that a part of the fresh-water smelt eggs 
that I had brought from Sidney [Lake Messalonskee, Maine] are in a fair way to hatch. The eggs 
are well developed and the fish lively, writhing, and lashing their tails. I should think about one 
in ten is alive. They were taken the evening of April 18th. 
In another place he stated that on May 5 the eggs were " infested with the white 
vegetable parasite,” but that it was not making much progress. 
It is not known how long the eggs obtained at Sidney on April 18 had been 
deposited, but from the time they were taken until the embryonic stage noted the 
time was about 17 days. However, another note, dated May 14, 1875, says: "Eggs 
from Brook-in-the-Woods, North Belgrade, hatched.” Computing from his notes, it 
would appear that it took either 30 or 45 days for them to hatch. 
In 1885 Fred Mather attempted to hatch salt-water smelts at Cold Spring Har- 
bor, Long Island, and subsequently published accounts of his experience. Before a 
meeting of the American Fisheries Society he read a paper on the subject, which was 
published in the “ Transactions ” of the society. It appears that his first effort was on 
March 4, 1885, when from a female fish that had been dead 15 minutes 30,000 eggs 
were taken on a bunch of coarse meadow grass and suspended in a glass tank with a 
flow of water from a Vs-inch cock, and in three days many were dead, and all died 
when one week old. 
On the 5th he repeated the experiment with eggs from a djdng female. In 5 
days three dead eggs showed, the sixth day 100 dead, seventh day one-fourth of the 
lot were dead. Up to the 17th, the thirteenth day after taking, there was little change, 
and on the 20th the eggs were put in a box outside the hatchery in swift water, as 
they began to show fungus. March 26 about one-half were alive, and these were in 
bunches covered by dead eggs and fungus. All the outside eggs were dead, and there 
was little hope of saving any. On April 3 the fish could be plainly seen in the lower 
eggs by removing the coating of dead eggs and fungus that had covered them for 2 
weeks. The eggs were again placed in the aquarium and 2,000 hatched on April 11, 
and on the 16th 9,000 more hatched and the rest were bad. About one-third of the 
eggs hatched under conditions that seemed hopeless and under which it would be 
impossible to hatch the eggs of salmon or trout. When the last eggs hatched the 
