FORMATION OF THE GERMINAL LAYERS IN TELEOSTEI. 213 



2. Of a thin film of cortical protoplasm entirely surrounding the yolk, and 



which frequently presents a considerable dilatation at the yolk pole. 



3. Of a number of filamentous protoplasmic processes, mainly confined to 



the base of the germinal area, which serve to keep up a communication 

 between the latter and the more purely nutritive yolk. 



4. Of the nutritive yolk itself, which constitutes the greater portion of the 



ovum. 



Thus it will be seen that the egg of the herring fulfils all the conditions of 

 Waldeyer's typical meroblastic ovum. The r61e played by these constituent 

 parts in the economy of the embryo is very marked. The part played by the 

 cortical protoplasm, and the root-like filaments, is particularly well brought out 

 in the herring. A discussion of the whole subject will, however, be deferred 

 until we consider the origin and growth of the parablast. Shortly before the 

 appearance of the first furrow the disc as seen in optical section has a diameter 

 of "84 mm., and is *28 mm. in thickness. On account of the collection of the 

 germinal protoplasm at one pole the egg loses its previously rounded outline, 

 and has now a diameter of 1*60 mm. in its longer axis, and 1*48 mm. in a 

 direction at right angles to this. 



Segmentation. 



The appearance and behaviour of the first segmentation nucleus in the fish 

 ovum has not been satisfactorily explained. Hoffmann (14), indeed, has 

 figured in a very diagrammatic manner the appearance and position of this 

 nucleus when it first divides; but so far as I am aware his observations have 

 not been confirmed, nor have they received any support from the work of recent 

 investigators. In the case of the herring I have used the most approved 

 methods of fixing and staining the material, but have as yet failed to observe a 

 nucleus of any kind until after the third furrow has been formed. The gradual 

 disappearance of the germinal vesicle in the ovum as it approaches maturity 

 has been already alluded to. Judging from analogy, a portion of the germinal 

 vesicle must remain as the female pronucleus. Having failed to demonstrate 

 this as a defined and recognisable mass, it appears necessary to assume that it 

 is distributed throughout the germinal protoplasm. I am not prepared to prove 

 this view, which undoubtedly is not in harmony with our information in other 

 cases, but it receives considerable support from a knowledge of the behaviour of 

 the germinal protoplasm during the early segmentation stages. At the time of 

 the appearance of the first furrow I have not succeeded in demonstrating a 

 nucleus either in the living egg or in prepared material. It has thus been 

 impossible to test the statements of Kupffer and Hoffmann as to the direc- 

 tion of the first plane of cleavage from actual observations of the nucleus, or of 



