METArHASE SPINDLE IN FOEFICULA AUEICULAKIA. 257 
'Coustrict ; in this case the distance between the centrosomes is 
found from fig. 73 to be 6*9 fx. Fig. 41 represents the equatorial 
plate at the conclusion of the inetaphase, and the dyads are 
seen to have become resolved into pairs of univalent spheres 
or rods, which are ready to move towards the poles. Fig. 74 
shows the centrosomes of this cell, and the length of the 
spindle, estimated from the known magnification, is 7*1 /x. 
Figs. 4*2 and 43 are drawings of the equatorial plate in the 
earliest anaphase, and the daughter-chromosomes are seen to 
be moving apart. The centrosomes of these cells are repre- 
sented in figs. 75 and 76 respectively, and their distances 
apart are 7*3 and 7*6 fx. It is difficult not to believe that figs. 39 
and 40 would have resembled fig*. 41, if fixation had not taken 
place until their spindle-poles were 7*1 ^ apart ; it is also diffi- 
cult not to believe that figs. 42 and 43 were identical with fig. 
41 at the moment when this was the distance between their 
poles. 
Let us now turn to cells of this generation in the testes of 
another specimen. Fig. 44 shows the equatorial plate before 
constriction of the chromosomes has begun ; figs. 45 and 46 
show the equatorial plate while constriction is in progress; and 
figs. 47 and 48 show it at the conclusion of the metaphase, 
when constriction is completed. The centrosomes of these five 
cells are represented in figs. 77 to 81 respectively, and their 
distances apart accord with those already observed for this 
mitosis; for the length of the spindle is found to be 6*6^t in 
fig. 44, 6*9 g in figs. 45 and 46, and 7*1 fx in figs. 47 and 48. 
In the circumstances I shall assume that the length of the 
spindle at the conclusion of the metaphase is a constant for 
the secondary spermatogonial mitosis, as we have already 
assumed it to be for the primary and secondary spermatocyte. 
The Ratios between the Lengths of the Mitotic Spindle 
AT THE Conclusion of the Sperhatogonial and Sperma- 
tocyte Metaphases. 
Having found that spindle lengths are probably constant 
at the conclusion of the metaphase of these mitoses, we 
