141 
Deaths, Tray 1 Tray 2 Tray 3 koh eae lgaloey 
March 24th............ 3 8 Ly 34 
hatched hatched hatched 
BA DOES. dis ce ry 1 6 140 
ae ae 4 93 81 68 
PACU s.cceecaees: 41 129 135 all dead 
PP AOU fovesaceacss 52 233 77 
We eeOUD. .. segs ccaave Unable to keep count any more. 
It will be seen from the foregoing, that the number of days 
these eggs were incubating in 1885-6 was more than in 1884-5 
(see On the Breeding of Salmonide, 1885), when they commenced 
hatching on the 82nd day, but not in large numbers until the 
87th. Inthe winter of 1885-6 (as will be observed in the table), 
the first young emerged on the 108rd day, but the main body 
did not begin to show themselves until the 106th day. In 
these two years the eggs came from Howietoun, were laid down 
to incubate in the same locality, and the identical hatching trays 
were employed, so doubtless variations in temperature may have 
been one of the causes. Mr Tyrer, F.M.S., has been good enough 
to furnish me with the following comparative statement for 
the two seasons, the thermometer standing in the open air. 
Season 1884-85 Season 1885-86 
Blecomburteae yt te he agry se 8) ee eie 
camropry ean oe. sae 36°6° Re 35°0° 
Bepruary” "9.075584 431° ape 33°6° 
LLG, AR A AMR 40-1° nite 39°6° 
Mean monthly temperature 39-9° ee 36°5° 
Although the mean temperature of the four months was in 
1885-6 3°4° below that of 1884-5, February, 1886, was nearly 
10° below that of 1885. It has been computed that the dif- 
ference in the time of incubation as related to temperature is 
_5 days to every degree, and this should have made the period 
in 1885-6 174 days longer whereas it was something over, or 
19 days. But another factor may here have had an influence, 
and that is the flow of the water. From the amount used it 
