116 W. R. B. ROBERTSON 
at one end and knowing the behavior of the unequal tetrads of 
other individuals (figs. 4-8), I believe that here we probably 
have the result of an unequal division of some preceding gener- 
ation where a No. 1 tetrad failed to divide at the proper place, 
in the middle between the two members, but instead so divided 
as to give a 14 portion of a No. 1 pair to one pole and a half- 
portion to the other pole (C; and C,). As a result of this, it 
might be imagined that there would be some generations of indi- 
viduals with a sesqui-valent (14-valent) No. 1 chromosome 
present pairing with the normal No. 1 (figs. D;, D.). The normal 
No. 1 might pair with the normal part, or with the fragment. 
In order to get a 24-valent No. 1 chromosome the parts of which 
are largely oriented in one direction it will now be necessary to 
suppose that, at some reduction division of the germ cells in 
succeeding generations of animals, a normal No. 1 chromosome 
became attached during the synapsis period to this sesqui-valent 
No. 1 chromosome. It might do this by becoming fused with, 
or by failing to separate from, the fragmentary (cut off) end of 
the sesqui-valent member after some method as is indicated in 
figures D, to Ds, giving as a result a 23-valent chromosome 
(fig. Ds), such as we have in figures 10 to 13. All portions of 
this chromosome, except the sesqui-valent fragment in the 
middle, would then be oriented in one direction (see arrows in 
figs. C,, Cs, and D,; to Ds). My reason for thinking that at least 
the terminal parts, which make up the greater part of the long 
chromosome are oriented in the same direction, is based upon 
the fact that I always see the chromosome arranged on the spin- 
dle in perfectly normal position, pointing directly toward the 
pole to which it is about to go, and upon the fact that the mantle 
fiber of this chromosome always springs from the longer (13) 
portion of the chromosome (figs. 12, 18, and B; to B;). The 
chromosome always travels with the larger, sesqui-valent 
portion in advance (figs. 12, 13, Bs and B;). 
In this animal evidently the normal No. 1 chromosome paired 
during synapsis with the distal No. 1 portion of the 23-valent 
chromosome (figs. 12, 13, and B, to B;). It should pair also 
