296 NATURAL SCIENCE. APRIL, 
other words, the orientation of the ovum is regulated by internal 
conditions. Born reached the same conclusion by another method. 
Having settled this point, Roux attacked the question whether 
all parts of the ovum were necessary to normal development. After 
two or more divisions, certain cells of the segmenting egg were pricked 
by a needle, so that some of the substance was lost. Thereafter, 
the development was watched until the embryo became differentiated. 
The result was that a loss of even one-sixth of the substance of the 
ovum produced only circumscribed local defects or disturbances. To 
develop fairly well, an ovum need not be intact. 
Then Roux proceeded to exclude from development one of the 
two first segmentation cells. This was done by puncturing one cell 
with a hot needle. The result was that a typical half-morula, or half- 
gastrula, or half-embryo developed. Thus, there might be half 
cerebral vesicles, one auditory vesicle, a half-gut, a single row of 
protovertebre, and a notochord of half the normal thickness. Some- 
times, however, these half-embryos were slightly abnormal. Thus it 
was proved that one of the first two segmentation cells may form half 
an embryo; that it has not only the requisite vital material, but the 
requisite power; and, that it can develop apart from its neighbour. 
It is a likely inference that in normal development, a similar indepen- 
dence does to a certain degree obtain. 
What is true when the first segmentation- “plane corresponds to 
the future median one, is also true when the first plane corresponds to 
the future transversal, as is sometimes the case in the segmentation of 
frog ova. In other words, Roux was able to produce not only a right 
and left half-embryo, but also an anterior and a posterior half-embryo. 
By destroying one of the first four segmentation-cells, he was able to 
rear three-quarter embryos; by destroying three of the four, he got 
quarter-gastrule. Indeed, he goes the length of maintaining that the 
gastrulation of the frog ovum is normally a kind of mosaic work, 
formed in at least four vertical, independently-developing pieces. 
Very remarkable is the process by which the half-embryo may 
form a whole by vitalising the injured half-egg which has been lying 
passive while the uninjured half has been developing. To this 
process Roux applies the term post-generation. Nucleiand, perhaps, 
also portions of protoplasm migrate from the uninjured side into the 
passive unsegmented mass; a remarkable kind of cell-division is 
set up; and gradually, in a peculiar manner which Roux describes, 
the missing half is post-generated. Quite lately, however, Roux has 
been able to rear an entire frog embryo from half an egg without any 
co-operation on the part of the other injured half. 
III. Professor C. Chun observed in 1877 that when the two first 
segmentation-cells of a Ctenophore ovum were shaken apart, each 
formed a half-larva, with four ciliated ridges, four meridional vessels, 
and one tentacle. The half-larve actually became sexual, and by a 
process of budding, the half awanting was formed. This observation, 
