two grated openings above, and passing out at two others astern, and
covered with a net to prevent escape. In a boat 13 or 14 feet long (on
the bottom) we put 10 or 15 salmon, to be towed a distance of 7 miles.
If the water is cool, twice as many can go safely, but there must be no
delay. It is very important that this car be smooth inside, with no
projections for the salmon to chafe on, and the gratings must be so
close that they cannot get their heads in between the bars.

If conveyance overland is necessary, a wooden tank 3 feet long, 2 feet
wide, and 2 feet deep, with a sliding cover, will take six salmon at a
time for a mile and perhaps farther, and they may be jolted along over
a rough road in comparative safety.

It has been our uniform experience that all the salmon that survive
till autumn were in normal condition as to their reproductive function,
and yielded healthy spawn and milt. On two occasions we suffered
serious losses of eggs. In neither instance could the loss be
attributed to any defect in the inclosure, but on one occasion the
conclusion was reached that the water which was well suited to the
maintenance of the fish was injurious to the eggs, rendering the shell
so soft that they could not be transported safely.

With the exception of the disasters enumerated above, there has been
but one that I can recall, and that was caused by the bursting of our
barriers at Dead Brook under the pressure of a flood.

BUCKSPORT, ME, April 7, 1884.