THE SMELTS 
335 
They were taken at 5.10 p. m., were all loose at 6.30 p. m., and at 7 p. m. next day 
many were stuck fast to the jar and the tubes. On March 30 those still loose were 
placed in another jar, and on April 2 a few dead ones were observed, while four days 
later the eggs grouped together in bunches that increased in size until on April 15 
the bunches were of the size of walnuts and covered with fungus. On the 20th a few 
hatched, and on the 21st all that were good came out. From this lot they got 60,000 
fish in 30 days with a temperature varying from 40 to 65° F. 
Concerning the fry, Mather said that they were the most minute of any that he 
had ever hatched and they were kept with difficulty. A strainer tube inclosing a 
a siphon such as was used for whitefish was entirely too large, for the fish passed 
through it with ease. After trying several things and having the aquarium overflow 
and the fish go out into the trout ponds he devised a spiral wire rolled on a stick of 
4 inches diameter and covered with thin muslin, which kept the fish and allowed a 
small stream to flow out of the siphon that was inserted. The lower end of this 
siphon was placed in a jar of water to prevent its going dry. It was said that the 
difficulty with siphon outlets was the tendency to empty faster than the inflow, 
consequently emptying themselves and then failing to start again, as they will suck 
no lower than the top of the jar holding the lower end. Of the eggs remaining 
attached to the first jar and its tubes in a single layer, not one hatched. Most of the 
fish came from eggs that were in masses surrounded by fungus. Mather said: 
This year’s experience upsets that of my eighteen previous years which taught me that the egg 
of a fish should be clean and free from fungus. I now except the smelt from the rule and think it 
[not] impossible that the embryo smelt must be protected from too much oxygen and good water by 
a coating of decayed eggs and fungus. 
In the discussion that followed the reading of Mather’s paper at the society 
meeting, H. J. Rice, among other things, said that according to his own experiments 
since 1876 and 1877 the result served to show greater success in hatching smelt in 
comparatively stagnant water than in any other manner. He said that the smelt 
appeared to be a peculiar form among fishes, and was no longer considered as one of 
the Salmonidse. According to Rice, young smelts will live in the same water for 
nine days, and fish-culturists would at once recognize the vast difference in this respect 
between these minute embryos and those of some of the Salmonidse, for which a con- 
stant change of water is absolutely necessary. The warmer the water the better 
the smelt appeared to thrive. He said that in the previous season and in the 
season in question experiments had been in progress to ascertain the feasibility of 
hatching young smelt in comparatively stagnant water and so far the attemps had 
been successful. Large numbers were hatched out with comparatively little trouble. 
In the same discussion Mr. Lyman said that he recollected that in 1867 and 
1868 attempts were made to hatch out some of the large variety called Belgrade (Me.) 
smelt. The eggs were put in somewhat swift running water in which trout eggs 
were kept, but none of them hatched. 
Mather (1894) read a paper at the twenty-third meeting of the American Fish- 
eries Society, entitled “Improved method of hatching smelts,” in which he reviewed 
previous work in smelt propogation. Mather said that outside of his own articles 
on smelt hatching in the fourteenth, fifteenth, and sixteenth reports of the society 
he could find nothing on the subject except an item in the paper of the late Prof. 
