738 Dig by. — The Somatic , Premeiotic , and 
some, but in the heterotype, prior to the formation of the bivalent chromo¬ 
some, the synaptic stage is intercalated. 
The linin becomes more and more confined to the nuclear periphery, and 
this causes an exaggerated appearance of parallelism as in the case of the 
somatic divisions (Fig. 37). Gradually the linin becomes contracted into 
more lumpy portions joined by the fine threads, which often run parallel to 
one another and crosswise (Figs. 38 A and 38 b). Once more the linin tends to 
become distributed throughout the nuclear cavity (Pis. LX and LXI,Figs. 39 A 
and 40). This is the definite beginning of the first contraction. The massing 
increases, the fine threads becoming obliterated in the general confusion 
though their double character is still marked (Figs. 39 A, 39 B, and 40). 
‘ Chromatic 5 bodies ( 3 ) may be extruded at this stage. The linin begins 
to collect at one side of the nucleus (PI. LXI, Fig. 41) until gradually the 
whole nuclear framework is drawn into the knot, which is balled together at 
one side of the nucleus (Fig. 43). Thus it seems clear that in Galtonia the 
parallel and sometimes fused linin portions which enter the synaptic knot 
as a general rule represent the concentrating sides of a single somatic 
chromosome, and not the approximation of two somatic chromosomes as 
«* 
described by Gregoire ( 10 ) (1907) and other workers in the heterotype pro¬ 
phases of animals and plants. As the parallelism found in the early 
heterotype prophases has the same appearance and the same origin as the 
parallelism found in the somatic prophases, it seems inconsistent to hold that 
in the heterotype prophases each parallel side represents a length of a whole 
somatic chromosome, and that in the somatic prophase each parallel side 
represents a length of half a somatic chromosome. It is true that there is 
so much variety in the thickness of the linin portions and so much general 
irregularity that it would be impossible to determine the significance of any 
specified parallelism. It is only by taking a broad and comparative view of 
the heterotype prophases in relation to the somatic prophases that one is 
forced to admit that the parallelism of the one is homologous with that of 
the other. Moreover, if the nuclei of the surrounding tapetal cells (PI. LX, 
Fig. 32) are compared with those of the pollen mother-cells there will be 
seen to be equally striking parallelisms present. 
If this interpretation of the parallelisms in the presynaptic stages of 
Galtonia be correct, then the presence of parallel threads in the prophases 
of parthenogenetic eggs would be explained ( 19 ) (1907). 
The parallelism in Galtonia is not so diagrammatically evident as 
figured in many plants. Though it may be said that the majority of the 
nuclear prophases of the pollen mother-cells do show parallelism to 
a greater or less degree, nevertheless nuclei can be frequently found in 
which the linin is of a more concentrated nature and no parallelism is 
apparent. 
