FORMATION OF THE GERMINAL LAYERS IN TELEOSTEI. 209 



fairly evenly distributed around it. The later measurements are taken at a 

 stage when the germinal protoplasm has collected into the form of a mound at 

 the animal pole of the ovum. The greatest diameter therefore passes through the 

 axis of the ovum at this stage. Whether the measurement 129 mm. represents 

 the largest diameter which is reached by the eggs of the Baltic herring I cannot 

 say, but it is probable that if there had been any further increase Kupffer 

 would have noted it. The early period at which this diameter was reached is 

 probably to be attributed to the temperature at which the experiments were 

 conducted. The temperature of the water during my experiments at the 

 Rothesay Aquarium varied between 40° and 42° F. ; whereas it is probable the 

 temperature was considerably over 55° F. during Kupffer's experiments, as 

 his embryos hatched out on the seventh clay. Dr Meyer (22) has shown that 

 young herring hatch out on the tenth or eleventh clay at a temperature between 

 51°*8 and 53°6 F. ; whereas at 32° F. the earliest embryos do not hatch until 

 the forty-seventh day. As the egg membrane leaves the yolk the surface of 

 the latter can be more easily studied. It is then seen that the " yolk granules " 

 on the surface are rapidly disappearing. Kupffer says they are dissolved. 

 However this may be, neighbouring "yolk granules" may be seen to run 

 together and flatten out into a thin pellicle, which soon becomes indistin- 

 guishable, with the large yolk spheres for a background. They behave, in fact, 

 very much like the small droplets of an oily nature which are found on the 

 surface of the egg of the cod before fertilisation. Whether the " yolk granules " 

 are really oil globules, or only yolk material richer in oil than the larger spheres, 

 I cannot say, but their behaviour would seem to support the former supposi- 

 tion. Kupffer next describes a series of clear vacuoles which arise at the 

 surface of the yolk as transparent spots. These increase rapidly in size, and 

 are pushed forward towards the centre of the yolk as a network of fine tubes. 

 With the appearance of the clear vacuoles the germinal protoplasm begins to 

 collect on the surface of the yolk. I have never been able to observe the 

 vacuoles which Kupffer describes, although I have searched for them 

 repeatedly. With the act of fertilisation an activity is set up in the germinal 

 protoplasm which causes it to collect rapidly on the surface of the yolk. Very 

 early in this process the outer yolk spheres are a little wider apart than those 

 towards the centre, and the protoplasm as it collects fills up the spaces between 

 them. A little later the protoplasm is almost entirely withdrawn from the 

 centre of the yolk, and there is then a thin layer of protoplasm on the surface, 

 with a number of branching root-like processes extending some distance into 

 the yolk. It is, I think, this collecting protoplasm which Kupffer has 

 mistaken for vacuoles and his series of coarse tubes. In sections of the egg 

 about this stage which have been mounted unstained the germinal protoplasm 

 is very transparent, whereas the yolk spheres are quite granular and of a 



