122 BULLETIN OF THE UNITED STATES FISH COMMISSION. 
placed in a similar box. At first very few died ; but on November 11 
there was a heavy rain-fall with a high southerly wind, which caused in- 
undation. The contents of one box were entirely destroyed, while 
those of the other suffered slightly by an overflow from the Qxara 
River. On December 24 the eyes of the embryos were first seen ; in 
the middle of February hatching began, and it was completed by the 
end of the month ; so the period of incubation lasted from one hundred 
and thirty to one hundred and forty days. Very few of the young fish 
died 5 and on April 29 all the fry, amounting to about 7,000, were 
planted in the Oxar& River, connected with the lake of Thingvallavatn. 
Most of the salmon eggs brought over from Reynivellir hatched, and 
the fry were deposited in the lake of Thingvallavatn on May 9. These 
were the first salmon in this lake. A year later one salmon was caught 
weighing one pound, but no others have been seen. 
The lake of Thingvallavatn is the largest in -Iceland, as it is from 4 
to 5 square miles* in area; and its northern part is 70 fathoms deep. 
The bottom consists partly of sand and gravel, but mostly of lava, as it 
is formed by volcanic eruptions. Around it is a great tract of lava and 
range of craters. Considerable animal life and vegetation are found in 
it. Trout (Sctlmo trutta) in this lake grow very large, up to 20 and 22 
pounds ; while mountain trout in several varieties reach the weight of 
from 1 to 7 pounds. Many trout are caught in this lake at all seasons ; 
during the winter numbers being hooked through the ice. 
Mr. Johanson, after superintending this hatching in the winter season 
of 18S4-’85, returned to Sweden. Since then the work at Reynivellir 
and Thingvellir has been kept up very poorly. In 1885- ? 86 about 6,000 
eggs were hatched at Reynivellir ; and in 1886-’87 somewhere between 
9,000 and 20,000 eggs were hatched. At Thingvellir a few thousand 
eggs have been hatched each season. A third attempt to establish a 
hatchery was made in the autumn of 1886 at Hjardarholt in Dalir, where 
several thousand eggs were hatched. All of these hatcheries are in- 
significant and badly managed. They have no suitable apparatus for 
such work, not a tank nor box being on hand for the transportation of 
fry or spawning fish. Last spring 2,000 salmon fry were brought from 
Reynivellir to Thingvellir in a pail. The young fish were said to be 
living on reaching Thingvellir, but they were not transferred to the 
Lake of Thingvallavatn that day, and being left in the hatching-trough 
over night, in the morning all were dead. The hatching-boxes are un- 
serviceable for hatching ; it is not possible to breed the fry successfully 
in such apparatus as there is ; and there are no basins for rearing the 
young fish. Consequently, the fry are turned out into the rivers and 
lakes before they have entirely consumed their yelk-sacks, and this 
probably results in the death or destruction of a great part of them, 
especially as sudden changes of temperature frequently occur in the 
water about this season from the presence of ice in the rivers. Then, 
* Danish square miles, which would make its area about. 100 English square miles. 
