DEVELOPMENT OP TEACHINTJS YIPEBA. 283 



saum of Kupffer), wliicli lies outside this, agaia widens also, 

 and often assumes a pointed form. Probably this widening is 

 the result of the invagination to form the auditory sacs ; but I 

 could not make out clearly what was going on until the process 

 was completed. The nasal pits appear also to originate about 

 this time. 



Kupffer's vesicle (postanal vesicle of Balfour) appears shortly 

 after the optic lobes are formed, soon after the growing blas- 

 toderm has passed the equator of the egg, and therefore a con- 

 siderable time before the closure of the blastopore. I have not 

 been able to detect the slightest relation between it and this 

 closure, although it certainly increases rapidly in size at the 

 period when the rim is nearly closed, and attains its maximum 

 development soon after the closure. It arises before any proto- 

 vertebrse are formed, and at the time of its disappearance there are 

 seventeen or eighteen somites. Its proximate origin is signalized, 

 as Kingsley and Conn state, by the appearance of a few granules 

 which draw together, and shortly afterwards the reside is seen 

 on optical transverse section as a very flattened lenticular body, 

 amber-tinted, and which, at its first appearance, seems solid. 

 Kingsley states that Kupffer's vesicle arises in the Gunner 

 when the blastoderm has covered over three quarters of the yolk, 

 and after many protovertebrae have been formed. This is con- 

 sequently at a much later stage in TracMniis. Eyder has, in 

 different species, noticed the first appearance of this vesicle at 

 periods varying from the time when the blastoderm covers three 

 quarters of the yolk up to nearly when the closure of the blasto- 

 pore takes place. In the Trout it appears when the blastoderm 

 has just passed the equator, and in the Perch it does not appear 

 until after the closure of the blastopore. 



About, or shortly after, the formation of protovertebrae, free 

 pigment-spots make their appearance, scattered irregularly over 

 the embryo. These increase in number and size until, at about the 

 time the heart begins to pulsate, they assume a stellate form. 

 They do not seem, however, to develop exactly pari passu with 

 the embryo ; at times they are slightly accelerated or retarded. 

 The assumption of the stellate form may also be either before or 

 after the heart begins to pulsate. Kingsley found that in the 

 Gunner they arise at the same time as the protovertebrae. 



The expanding blastoderm continues to grow over the yolk, 

 until gradually it leaves a mere pore enclosed by the thickened 



