334 
BULLETIN OF THE BUREAU OF FISHERIES 
mass of dead eggs was rotten and foul. The temperature ranged from 40 to 42° F. 
In taking the eggs the grass was laid in a milk pan and covered with water. The 
female was manipulated first, and as the eggs do not stick fast until some minutes 
after being taken, perhaps after impregnation takes place, they were distributed 
evenly over the grass with the tail of a fish. 
Again Mather (1885) said that, knowing nothing of smelt hatching, the litera- 
ture of which was meager, he determined to try several plans. So, on March 5, 
50,000 eggs were taken from a weak female, on stones the size of a man’s fist, in 
water, and placed outside the building in a covered waste trough, which took the 
water from the house to the ponds. The current was slow but the eggs washed off 
refusing to stick in bunches as they did on the grass. The consequence was that 
the stones were covered with eggs only one layer deep. Three days after this they 
looked well, but in a week were all dead, though no fungus had formed. 
He tried again on March 8, by taking about 70,000 eggs by the dry method on 
tiles, letting them stand five minutes before adding water, and then placing them in 
one of the hatching troughs. On the 16th one-half were dead, and on the 24th they 
were covered with fungus. On April 7 there had been no change, the eggs under- 
neath the fungus were bright and good, but they were left unattended until the 12th, 
when the trough was found empty. The other attendants pronounced them dead 
and threw them away. Mather did not feel certain that they were all dead, for 
his experience that year told him that it required an expert to judge of this. He 
said that a mass of smelt eggs all rotten on the outside and covered with fungus 
half an inch thick should be given the benefit of all doubt and be carefully examined 
before condemnation. 
Another trial was made on March 9, when 100 more fish that had been taken 
in seines were obtained. The first lot were so badly injured by gill nets that they 
were covered with fungus in a few days. On the 12th 70,000 eggs were taken on 
tiles and stones in water and placed in a trough that received the flow from nine 
hatching troughs, and consequently carried a swift current. These eggs were evenly 
distributed over the tiles and stones several deep, and did not flow off as in previous 
cases. Not until March 22, 11 days after, were any dead eggs or fungus seen. At 
5 days old the formation of the embryo could be seen by means of a microscope, 
and at 15 days the fish could be seen with the unassisted eye. At this time fungus 
had spread all over the outside eggs, but underneath there were but few dead ones. 
On April 6, when the eggs were 26 days old, they were placed in the glass tanks 
with a flow from above and a siphon outlet, and 4 days later began hatching fast; 
2 days later there were 11,000 fry, all that were obtained. The temperature varied 
from 37 to 58 degrees, and the time of hatching was 30 days. The water used in 
all these experiments was pure spring water. 
The last trial was in the McDonald hatching jars and was the best of all, pro- 
ducing 60,000 fish from 200,000 eggs. They were taken on March 21 by the dry 
method, let stand five minutes, half a pint of water added, and kept in motion 20 
minutes by tipping the pan from side to side and occasionally using the tail of a fish. 
The object of this was to keep the eggs from sticking together, so that they might 
be treated as free eggs. After this more water was added and the eggs allowed to 
rest for 20 minutes. They were then washed twice and placed in a McDonald jar. 
