TWELFTH ANNUAL MEETING. 39 
tuted, in the basement of the Smithsonian Institution, a series of 
experiments in order to determine the limit of healthy retarda- 
tion of development that could be effected by lowering the tem- 
perature of the water employed. In order to subject the eggs 
conveniently to the action of the current of cold water, they 
were placed in small two-ounce laboratory flasks, closely corked. 
Through the centre of the cork was passed a glass tube which 
descended to within a short distance of the bottom of the flask, 
and through which the current of water was admitted to the 
apparatus. This is shown in fig. 3. 
An exit tube, the lower extremity of which extends a short 
distance below the neck of the bottle, provided for the escape of 
the water. Whilst this form of apparatus had been devised by 
me in connection with the experiments on retardation above re- 
ferred to, I had no sooner fixed upon the apparatus than I felt 
at once | had arrived at the solution of the question of automatic 
separation of the dead from the living eggs. An eight-ounce 
wide-mouthed glass jar, such as is used in the National Museum 
for holding alcoholic specimens, was fitted up as indicated 
(fig. 3). 
Six thousand shad eggs were placed in this apparatus and a 
current of water turned on and regulated. The movement of 
the current established a regular rolling, boiling motion on the 
eggs, which brought all in succession to the surface. The dead 
eggs remained there, forming as they were freed from the mass 
a layer upon the upper surface of the others. By pushing down 
the exit tube a suitable distance, I found that the dead eggs 
were taken up by the escaping current—were by degrees drifted 
under the lower end of the tube, lifted through it by the current, 
and swept out, leaving an absolutely clean mass of live eggs in 
the jar. 
This lot of eggs was successfully hatched, and at the time of 
hatching not a dead egg was found in the bottle, nor do I think 
a live egg was lost in the whole course of the experiment. 
The first experiments had been framed solely with reference 
to the assumed slight difference in the specific gravity of the 
living and the dead eggs. Attentive study of the movement,of 
the eggs in the jar showed a still more potent influence for 
