1892.] Experimental Embryology. 591 
In this way as many as fifty cells were isolated from the two 
celled stages and kept separate in small vessels in which they 
could be examined microscopically and actually seen to 
develop. The separated cells develop at first, as do the unin- 
jured cells of the frog in Roux’s experiments, but the ultimate 
result is that each cell (of the two after the first cleavage of 
egg) may give rise to a complete though small larva. 
This is found to be the case in both Echinus and Spheer- 
echinus. 
In the normal cleavage of Echinus, according to Selenka, 
there are formed two, four, eight cells, and then four at one 
pole bud off four small cells, while the four at the other pole 
divide equally. Now in the cleavage of the half egg Driesch 
finds two and then four cells, followed by an eight celled stage 
which is formed by or budded from two little cells at one pole, 
as opposed to an equal division of the two cells at the other 
pole; thus the eight celled stage is exactly the half, in form 
and arrangement, of the normal sixteen celled stage. 
These half formations, however, become over night con- 
verted into complete blastulas having half the normal size, but 
apparently made up of cells of normal dimensions, so that we 
may infer there is half the normal number of cells. About 
thirty dwarf blastulas were obtained from isolated cells, and 
from these normally formed gastrulas, and eventually, in three 
cases normally formed dwarf plutei were reared. Thus it 
would be possible to obtain two plutei from one egg by sepa- 
rating the first two cells of cleavage. 
The application of this to the explanation of the occurrence 
of twins is made easier by the actual finding of numerous 
abnormal stages in cleavage and gastrulation resulting in the 
production of twin gastrule or larve, or in some cases of com- 
bination of three-fourths and one-fourth blastule. As these 
occur so frequently in material submitted to the shaking pro- 
cess, which variously effect different eggs, we have reason to 
suppose the twins are directly due to the mechanical separa- 
tion or disturbance of the material in the egg. 
The heterogeneous character of the various experiments 
referred to in the present article and the conflicting and often 
