Four days after the artificial fertilization eggs of the pond smelt, Hypomesus olidus PALLAS, were puttin 20 rials with water. Two of them were then placed in each of 10 compartments (Nos. 1-10) of a SENÔ-TAUTI'S serial incubator. They ran at temperatures differing from each other by nearly 1•5 degrees, No. 1 at the highest (19•9C) and No. 10 at the lowest (6•4°C.) With a view to determine the effects of temperature variations on the developing eggs one of the two vials in each compartment was interchan-;ed with the corresponding one in another, but the other vial in respective section was kept in situ as control. The interchange was made in 5 p.sirs, viz., No. 1 with No. 10, No. 2 with No. 9, No. 3 with No. 8, No. 4 with No. 7, and No. 5 with No. 6. The results obtained may be summarized as follows:-
1) Water temperature for normal development of the eggs ranges from 19° to 7°C.
2) The relation between the temperature (θ) and time in days (T) required for the deve- lopment of eggs may be expressed, as given by HIGURASHI and TAUTI, by the formula
Te10=constant,
in which a=0•138, The. temperature coefficient for the embryonic development of the pond smelt was found to be Q10=3•98.
3) Within the range of mean deviations 6•8°-0•6°C. in the series of temperatures 19•9°-6•4°C., the larger the temperature deviation the less the hatch-rate of the fry and the less the, time required for hatching.
4) TAUTI'S formulae concerning the relation between the mean deviations of the water tem-perature (M) and time in days (TM) and that between the former and the percentage of the hat lied fry((N-NO)M), viz.,
TO/TM-1= a2/2M2, and (N/N0)M=(N/N0)0 e- ??
where T0 and (N/N0)0 are the time and the percentage when M=0 respectively, hold true for the writer's material with the values of the constants a=0•115, c=0•017, Q10=3•16.