x 
120 W. R. B. ROBERTSON 
No. 4 to which it corresponds, or with which it is homologous, 
and leaving the distal end of the normal No. 4 projecting beyond 
(fig. F:), would, on fusion with it, form a single cylindrical body 
having one end, the distal, smaller in diameter than the remainder 
of the cylinder. If a splitting occur at the end of synapsis in a 
plane formed regardless of the old plane of fusion, having merely 
for its object the splitting of this single fused body into sym- 
metrical halves, there would result two daughter chromosomes 
of equal size but with smaller diameter at the distal end than 
throughout the remainder of the chromosome (figs. F; to F;). 
There would then be no such uniform inequality of the No. 4 
chromosomes preserved in the first spermatocyte divisions of 
this animal as we have found. © Instead of having a normal sized 
No. 4 and a (say) five-sixth-sized No. 4 in every cell after the 
first maturation division, there would be two No. 4 chromosomes 
of equal length but shorter than the normal (fig. F;). On the 
contrary every first spermatocyte metaphase and anaphase 
(figs. 5-7) shows one normal sized No. 4 chromosome of the same. 
fixed relative size and likewise one deficient No. 4 chromosome 
of a definitely fixed relative size. 
If we turn to the abnormally long tetrad (figs. 12, 13), we have 
even more definite proof that fusion longitudinally and splitting 
along a new plane in all probability does not occur. The normal 
No. 1 chromosome evidently has paired with the distal portion 
of the long No. 1 chromosome in a manner similar to the diagram 
shown in figures B,; and By. If in such pairing it fuse completely 
with its longer mate, its would form a cylinder again, with one end 
of larger diameter, the opposite of smaller diameter (figs. G, to 
G.). If, on the splitting of the fused thread, the new split be 
formed, not along the old plane or fusion, but upon any plane 
giving two equal daughter threads, we would get two long chro- 
mosomes of the same length with diameters large at the distal 
end instead of a long chromosome and short chromosome, of the 
same relative lengths and equal diameter, with which we started. 
Again, if the new split appeared in the fused chromosome in a 
plane at right angles to that of fusion (figs. H, to Hs), we should 
~ 
