SEGMENTATION OF THE OVUM. 9 



vertical and parallel to the first furrow ; it divides each of 

 the four segments into two parts (fig. 7). Meanwhile the 

 larger, clearer masses of the opposite side of the ovuna are 

 gradually separating from the darker cells of the animal pole, 

 and breaking up into smaller masses of various sizes and irre- 

 gular shape; they will henceforth be called the endoderm 

 masses or cells, inasmuch as they give rise to the greater 

 part of the endoderm. The small, darker cells, each of 

 which contains a central clear area — the nucleus, give rise to 

 and will be called the ectoderm. 



Each of the ectoderm cells, which are apparently eight in 

 number and arranged in two rows (four in each row), now 

 divides along a line parallel to the long axis of the ovum, so 

 that there are sixteen ectoderm cells arranged in four rows with 

 four cells in each row. 



The ectoderm cells continue to divide in this manner, and 

 the endoderm masses become more broken up, until at the 

 end of segmentation (fig. 8) the ovum consists of a number 

 of large and small branched endoderm masses scattered 

 irregularly within the egg membrane, while the ectoderm 

 cells consist of a mosaic of more or less hexagonal cells closely 

 applied together and placed close to the membrane on one side 

 at about the middle of the long axis of the egg. 



The egg at this stage presents a very peculiar appearance, 

 and I would not believe for some time that I was not dealing 

 with an abnormal or injured ovum. But I found the stage so 

 often, and so many stages intermediate between it and the 

 earlier and later stages of development, that I cannot but believe 

 in its normal existence. I found it also when every precaution 

 was taken to avoid injuring the ovum ; when I merely opened 

 the animal and examined the ovum through the transparent 

 walls of the oviduct without even touching any part of the 

 female organs. 



The endoderm masses at this stage — and I have no doubt this 

 is the case in other stages, but in this case the fact can be 

 clearly seen — are branched, and the branches of adjoin- 

 ing masses in some cases anastomose. One must sup- 



