FORMATION OF THE GERMINAL LAYERS IN TELEOSTEL 235 



blastic ova. As already pointed out, the vegetative pole in a Teleostean such as 

 the herring consists at first of only one cell, or if it contains more than one 

 nucleus the protoplasm is at any rate unsegmented. The protoplasm is mainly 

 peripheral, and surrounds a practically solid mass of passive yolk. There are at 

 most a number of protoplasmic filaments pressing down amongst the yolk 

 spherules which serve to bring the active and passive material into closer union. 

 Too much stress cannot be laid on this fact, for to my mind it constitutes the 

 main difference between meroblastic and holoblastic ova. A certain proportion 

 of yolk to protoplasm may or may not prevent total segmentation, the result 

 depending on their relative distribution in the ovum. 



The development of the Decapod Crustacea shows that, to commence with, 

 total segmentation may take place, and then that later a central unsegmented 

 yolk mass may be formed, while the protoplasm and nuclei accumulate on the 

 surface. This appears best explained by supposing that the protoplasm, when 

 generally distributed throughout the yolk, was present in sufficient quantity to 

 bring about total segmentation, but that, as it collects at the surface, a central 

 mass of practically pure yolk is formed, which can no longer be assimilated in 

 the same manner as that in a typical holoblastic egg such as that of Amphioxus. 

 In this manner an ovum at first holoblastic becomes secondarily meroblastic. 

 To return to our Teleostean ovum. The protoplasm in the vegetative pole 

 increases rapidly in bulk by an assimilation of its enclosed food material, and 

 thus is enabled to bud off cells which, had the distribution of yolk and 

 protoplasm been otherwise, would have been produced by normal segmentation. 

 Thus arises the distinction between primary and secondary segmentation, and 

 the latter is seen to be only a modified form of the former. The first equatorial 

 furrow, whenever it arises, divides the animal from the vegetative pole, and in 

 meroblastic ova the segmentation in the vegetative pole, by becoming of the 

 secondary type, accommodates itself to any relative proportion of yolk. 



According to my observations, the separation of the animal from the 

 vegetative pole in the herring occurs with the formation of the third furrow. 

 It may be, however, that this is not the case in all other Teleosteans. I never- 

 theless regard the furrow or partial furrow which divides the peripheral proto- 

 plasm from that which undergoes primary segmentation in the germinal disc, as 

 the equivalent of the first equatorial furrow in holoblastic types. 



The archiblast in the herring, together with the cells derived from the 

 parablast, prior to the formation of the segmentation-cavity, give rise to the 

 epiblast. The vegetative pole then gives rise to the primitive hypoblast, which 

 is in turn differentiated into the mesoblast and permanent hypoblast. I am 

 not at present prepared to say whether this is the case in most of the 

 Teleosteans. It appears, however, on a priori grounds, that at any rate in 

 those forms which have a vitelline circulation the process may be modified. It 



