962 THE AMERICAN NATURALIST. [Vor. XXXV. 
each piece soon closes, and a new sphere of smaller size forms 
from each piece. These spheres then proceed to pass through 
the later stages of development characteristic of the sea-urchin 
development. Driesch has found that if the embryo is cut in 
two at a later stage, when the process of gastrulation has 
begun, not only the piece that contains the part about to 
invaginate will continue to develop and produce a whole embryo, 
but the piece from the opposite side will also form a gastrula 
and embryo. If, however, the same experiment is carried out 
after the gastrulation has been finished, these pieces of the 
outer wall fail to produce a new gastrula. A similar result 
was obtained in still later stages in regard to the formation of 
the two pouches that pinch off from the inner end of the 
archenteron of the starfish embryo. If the inner end of the 
archenteron is cut off before the two pouches have formed 
there, they will be formed again from any part of the more 
proximal portion of the archenteron; but if we wait until the 
pouches have once been completely formed by the archenteron, 
and then cut off the inner end of the latter, the two pouches 
are not made anew at the cut-end. In other words, after 
the archenteron has once produced the two pouches it seems 
to lose ¢shroughout its entire length the power of repeating the 
process, although at an earlier stage all the parts possessed 
this power. 
Another somewhat similar result has been recently obtairied 
by Spemann. If a thread be tied around the middle of the 
embryo of the frog at the time when the medullary folds are 
appearing, the egg being partially constricted by the thread 
into a more anterior and a more posterior part, the latter will 
produce at its anterior end (where the string constricts the 
nerve plate) a new head, and there is produced in this way 
a double embryo with two heads, one at the anterior end, and 
one behind this at the middle where the constriction is present. 
The result shows that the material of the dorsal nerve plate is 
totipotent in so far as the formation of all the structures of the 
brain, eyes, and nerve cord are concerned, and that the position 
of any part in relation to the rest determines its differentiation. 
The brain forms at the anterior end of the nerve-plate, but it 
