EMBRYONIC DEVELOPMENT OF THE TELEOSTEANS. 47 



rests by its plane face on the intermediate layer with which it is 

 seen to be everywhere in immediate contact. It is composed of 

 polyhedric cells which are very clear and in each of which it is 

 easy to distinguish, even in the living state, a large, spherical or 

 slightly eiUpsoidal nucleus. The cells are not, however, pressed 

 one agaiust the other in every part : here and there among them 

 small spaces of varying form and dimensions can be observed. 

 But neither at this stage of development nor at any moment 

 during segmentation does there exist within the thickness of the 

 disc any ' segmentation cavity.' It is well known that this 

 cavity, discovered first of all by Lereboullet (4), in the Perch and 

 the Pike, was re-discovered by Van Bambeke in the common 

 Eoach. The superficial cells are polyhedric as are the subjacent 

 ones and also those which limit the blastodisc inferiorly. The 

 cleavage mass or blastodisc does not therefore at this moment 

 exhibit any trace of delamination. The only difference which 

 the cells of the disc present is relative to their dimensions; the 

 volume of the cells increases somewhat in passing from the 

 surface to the deeper parts. 



The intermediate layer has undergone important modifications. 

 It is possible now to ascertain the existence, throughout the 

 extent of this layer, of a great number of nuclei, generally- 

 oval, with very sharp contours and provided with one sometimes 

 with two punctiform nucleoli. All these nuclei have very nearly 

 the same dimensions, they are a httle smaller than those of the 

 cells of the blastodisc. If the focus of the microscope is 

 arranged so as to give a surface-view of the deutoplasmic globe 

 and if the region which immediately fringes the blastodisc be 

 examined, a finely granular zone is distinguished there in which 

 nuclei disposed regularly and at equal distances are scattered. 

 It is impossible to make out any delimitation of cell-areas, but 

 around each nucleus a small zone of a more granular character 

 is seen in which a very manifest radial striationis apparent. The 

 latter becomes much more obvious after the action of 1 per 

 cent, solution of acetic acid. The perinuclear zones are sepa- 

 rated one from another by clear spaces devoid of all granulation. 

 These spaces together form a network in the meshes of which 

 the nuclei with their radial haloes are disposed. It is clear that 

 it is this part of the intermediate layer, namely, the part 

 situated outside the blastodisc, which has been observed by 

 Kupffer (5) in Sjjlnachia and Gasterosteus, and which has re- 

 ceived from this excellent observer the name of "Koinerzone" 

 (nuclear zone). I have also been struck with the remarkable 

 resemblance between the objects which I have just described and 

 that cell-layer which appears in the Cephalopoda at the surface of 

 the. deutoplasm and is formed by the elements discovered and de- 



