214 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
With the scaff-net transfer a portion of the eggs from the pan to the floating tray. A 
slight sifting motion of the tray will cause any remaining trash to fall through the 
meshes, and, at the same time, level oft" the eggs. The dead are picked out with the 
tweezers and the living are put in a jar. By counting the number of trays or layers 
the proportion of the entire lot can be ascertained for each tray or layer. Knowing this, 
it can approximately be determined what quantity each jar is to receive. With heavy 
eggs it is not advisable to fill the jar more than one-third full, though excellent results 
have been obtained with the jar three fourths filled with Atlantic salmon eggs. With 
heavy eggs the boiling motion is neither necessary nor desirable. The amount of 
water for each jar is about the same as that required for shad, possibly a little less, 
and it can be used under much less pressure than that needed for semi-buoyant eggs. 
No motion being given to the heavy eggs, there is of course no automatic separation 
and picking of the dead eggs. To remove the dead eggs the jars have to be unsealed, 
the eggs poured into a pan and picked with tweezers in the same manner as when they 
arrived. This slight trouble, however, is abundantly compensated for, because in this 
apparatus the heavy eggs do not rt quire picking for several days. So perfectly is 
every egg supplied with the constantly-changing envelope of fresh water that they 
remain healthy for a much longer period than in the forms of apparatus employing 
horizontal currents. In the jar there are no dead angles for stale and eddy water. 
Should tlie water become muddy, as it does at times in nearly all hatcheries, eggs 
exposed under horizontal cnrrents are in constant danger of asphyxiation from deposits 
of sediment. In the McDonald jar, the water coming from below and passing upwards, 
all sediment deposited is on the upper side of the egg. Gravity prevents any accumu- 
lation on the under side, which always presents a clean surface to the incoming current. 
The time elapsing between the “ pickings,” even with a uniform temperature, will vary ; 
and here again the attendant must be guided by close observation and judgment. Mr. 
Charles G. Atkins says that in water of 40° Fahr. the eggs should be picked once in every 
three days. This rule was established with the old trough-aud-tray system employing 
horizontal currents, and was undoubtedly correct for that system. During the season 
of 1883-84, at Central Station, in Washington, D. C., ajar containing 30,000 Lake trout 
eggs remained closed from January 6 to January 21, when it was opened more for the 
purpose of washing than picking, but no unusual mortality could be detected. The 
temperature of the Potomac Eiver water during that time varied from 41° to 38° 
Fahr., and averaged 39.5° Fahr. It may be interesting to note that during this period 
the water was uunsnally muddy. As often as is thought desirable the sediment may 
be washed from the eggs without opening the jar. Grasp the jar firmly at the bottom 
and top, lift clear of 4he table, and turn it rapidly so that the sides may generate a 
double cone, the center of the jar forming a common apex. In this operation care 
should be taken that the turning of the jar is not so violent as to throw the eggs up to 
a height that they will pass ofi' through the outlet -tube. The eggs rolling upon 
themselves cut off the sediment, which readily passes off with the waste water. This 
operation will cleanse the eggs tboroughlj", and, if judiciously repeated, will obviate 
all danger of asphyxiation by deposit of sediment. During the time when the fry are 
breaking from the shells the jar should be opened somewhat oftener, the dead picked 
out, shells removed, and the fry transferred to proper rearing quarters. The eggs 
should now be handled with more care than formerly, for while a salmon or trout egg 
is quite tough and will stand a good deal of rough handling at certain periods, as the 
