228 MR GEORGE BROOK ON THE 



this process of cell formation in the parablast, and each nucleus arises inde- 

 pendently of its neighbour, in a manner similar to that which I have described 

 for Track inns. The observations of Kupffer, Klein, and others, are very clear 

 on this point ; and I have frequently observed the full formation of nuclei both 

 in the living Qgg and in the prepared material. The protoplasm in which these 

 nuclei appear is, however, part of the original germinal layer of the ovum ; and 

 it thus appears probable that they are to be regarded as derivatives of the first 

 segmentation nucleus. I thus regard the cells formed in the parablast as 

 secondary segmentation products in the sense of Waldeyer, and must leave 

 the question of nuclei open until we have more detailed information on the 

 subject. If, indeed, the observations of Hoffmann should prove true for all 

 Teleostean fishes, the question presents no further difficulty ; but, as already 

 stated, I have not been able to accept Hoffmann's views. 



The' cells thus formed in the parablast are next set free from their bed of 

 unsegmented protoplasm, and join those derived from primary segmentation 

 in the archiblast. About the same time the cells in the archiblast undergo 

 division, and soon the cells derived from the parablast are no longer distin- 

 guishable from the archiblast cells. As already stated, fig. 12 represents a 

 section of an ovum of the herring twenty-six hours after fertilisation. In 

 fig. 14, which is from material preserved two hours later, the cells derived 

 from the archiblast are easily distinguished from those which have been 

 recently added from the parablast. The archiblast cells stain more deeply, 

 and besides nuclei there are indications of an intracellular reticulum. Perhaps 

 the most noticeable characteristic is that the archiblast cells are loosely col- 

 lected together, and present in section a number of vacuoles between adjoining 

 cells which have not yet become completely separated. The parablast cells, on 

 the other hand, only stain faintly, show little structure, and are closely crowded 

 together beneath the others. In fig. 1(3, which represents the appearance two 

 hours later again, all difference between the two sets of cells is lost, and if it 

 had not been for the two previous stages, one would not have known that the 

 archiblast had received any addition of cellular elements from the parablast. 

 The unsegmented portion of the parablast still remains as a somewhat thin 

 film beneath the segmented blastoderm, and again increases rapidly in bulk by 

 assimilation of food yolk. Four hours later than fig. 15 a number of cells are 

 again in process of formation, which are destined to be included in the seg- 

 mented blastoderm in a similar manner to the first batch. Fig. 17 represents a 

 section of this stage. The basal jiortion of the segmented blastoderm is repre- 

 sented with its adjoining parablast. The two portions are not in contact at the 

 periphery, but this is the result of a mechanical injury. An endeavour has 

 been made to represent as nearly as possible the exact appearance and struc- 

 ture of each cell in the portion of the section represented. A gradual transi- 



