56 
it is found in practice to produce no increase in the num- 
ber of eggs saved. 
During the present season Mr. Dwight Lydell, an em- 
ployee of the Michigan Fish Commission, has kindly made 
for me some tests of this question, the results of which 
are as follows: 
On May 13 about five and one-half million eggs were 
taken by the starch method at Roberts Landing, on the 
St. Clair River: these were shipped by steamer to Detroit 
and placed in closed jars. On May 24 the good eggs re- 
maining in the jars were measured, and it was found to be 
35 per cent. of the number placed in the jars. Eggs not 
treated by this method showed 30 per cent. of good eggs. 
The advantage of 5 per cent. thus apparently gained by 
the use of the starch is so slight as to be scarcely worth 
consideration. 
It then occurred to me that perhaps the agitation due to 
shipping the eggs, and to the repeated handling in the 
hatchery in order to remove injured eggs, might be respon- 
sible for the loss. The eggs are usually shipped on the 
day following that on which they are taken. At this time 
the material which is to form the embryo lies on one side 
of the yolk in the form ofa little disc or cap, the germinal. 
disc. The rest of the yolk is covered by only a very thin 
layer of protoplasm, which is continuous with the germinal 
disc at its edge. 
This layer of protoplasm is very easily ruptured so that 
the yolk protrudes through it into the space between the 
yolk and the shell. The yolk which thus escapes turns 
white and the egg dies. Early in the history of the egg 
the germinal disc extends itself over the yolk, and finally 
completely covers it. The yolk is thus protruded by a 
firm layer of much greater thickness than the original 
protoplastic covering. This layer is not easily ruptured, 
and the egg is consequently less liable to injury by me- 
chanical means. 
