Cardiff : Synapsis and reduction 291 



Whether the one large nucleolus resulted from the union of sev- 

 eral smaller ones could not be ascertained, though this seems 

 probable. Around this nucleolus the chromatin threads show a 

 marked tendency to collect. It was also noted, in other forms as 

 well as Ginkgo, that the moieties are in contact, or at least closer 

 together, in the vicinity of the nucleolus (figure 54). The con- 

 traction of the chromatin continues until a dense knot is formed 

 at one side of the nucleus (figures 55 and 58). In stages just 

 previous to the close knot (figure 55) and in those following, there 

 is very rarely present a nucleolus in Ginkgo. This is a marked 

 difference from the nucleolar behavior in other forms and 

 indicates that the nucleolus may be playing a .more promi- 

 nent part in chromatin formation than generally supposed. 

 This unusual behavior of the nucleolus may be correlated in 

 some way with the other unusual cytological conditions in 

 Ginkgo. 



As the chromatin threads extend from the synaptic knot they 

 nearly always show their paired character, especially if the closely 

 contracted stage is not yet reached, as in figure 55. In these por- 

 tions of the threads that are still free from the knot the chromomeres 

 are clearly discernible (figure 55), while after they have been 

 drawn closely into the knot the threads appear as a continuous 



chromatin-mass (figures 56 and 57). 



As the threads emerge from synapsis they are much shorter, 

 thicker and more homogeneous throughout than in the presyn- 

 aptic stages (figure 59). 



Their bivalent character can be made out with difficulty until 

 the chromosomes are finally formed in prophase when it is quite 

 evident (figure 60). The twelve chromosomes figured here are 

 from a camera drawing of an unusually good early metaphase 

 view. While this number may not prove correct, it corresponds 



with a number of other counts. 



Owing to the fact that a blepharoplast is formed in the male 

 gametes of Ginkgo, it was thought worth while to make a search 

 for centrosomes in the reduction divisions where they would prob- 

 ably be in evidence if they exist at all. No structure of any kind 

 which could be interpreted as a centrosome could be found, the 

 spindles being very short and thick. 



