734 
otherwise normal (fig. 1, C). (2) If (as frequently happens) the 4-celled 
stage falls into two pairs of cells, each pair may give rise to an 
embryo of half the normal size (proved by isolation-cultures). (3) If 
the four blastomeres be imperfectly separated three types of multiple 
embryos (gastrulas) arise, namely: (a) double embryos (each half 
presumably arising from a pair of cells; (b) triple embryos, one 
body being twice the size of the other two; and (c) quadruple 
embryos, each body being one fourth the normal size. 
The most important and interesting results were obtained by 
experiments with the 8-celled stage. Although my stay at Faro could 
not be sufficiently prolonged to admit of an exhaustive investigation, 
the following result is I believe established with a considerable degree 
of probability, though not with absolute certainty, viz: The isolated 
blastomere of the 8-celled stage is incapable of pro- 
ducing a gastrula. The normal 8-celled embryo (fig. 2, A) con- 
sists of four “macromeres” and four slightly smaller “micromeres”, and 
there in reason to believe that the micromeres contain all, or nearly 
all, of the ectoblast. These cells may easily be shaken apart and 
isolated in the hanging drop or otherwise, a normal embryo being 
placed in each culture as a control. After undergoing a cleavage at 
first closely similar to that of an entire ovum, the isolated blastomeres 
give rise to a great variety of forms most of which acquire cilia, swim 
actively about, and may live until the end of the second day or even 
longer. Among them (apart from monstrous and irregular forms) are 
three principal varieties, viz: (a) perfectly flat plates of cells, (b) 
more or less curved plates, and (c) blastulas one eighth the normal 
size, either closed or with a pore at one side. Of these three forms 
the closed blastulas live longest and swim most actively. The curved 
plates usually appear under two forms in one of which the plate is 
only about half the thickness of the other, and the two forms closely 
resemble fragments of the normal ectoblast and entoblast respectively. 
In the formation of the plate-shaped embryos the + blastomere seems 
in fact to undergo in the main the same development it would have 
had if still forming a part of a normal embryo — i.e., a partial 
development — giving rise to a partial embryo incapable of regene- 
rating the missing portions, and hence incapable of complete deve- 
lopment. The 4 blastulas are less easy to understand and require 
further investigation, but isolation-experiments seem to show them as 
incapable of further development as the plate-forms. 
None of the 4 embryos, as I believe, are capable of full deve- 
lopment. I have isolated a considerable number of the 5 blastomeres 
