58 Papers from the Marine Biological Laboratory at Tortugas. 
Cases of undoubted abortive attempts (fig. 74) at a division were seen, but 
never a consummation of such process. The isolated case, therefore, must 
stand as an exception to the ordinary process of maturation. 
The second polar body is constricted off in about 30 minutes after the 
first is formed. The process of spindle formation must here be very rapid. 
As far as I know, the process of spindle formation between the two centro- 
somes of the central pole of the first spindle has never been described in 
echinoderms. Miss Hogue figures a second maturation spindle in tangen- 
tial position and believes she has evidence that the spindle is always so formed 
and subsequently revolves into radial position (p. 525). Mathews (1895) 
states that “the outer centrosome of the second polar spindle is formed at 
the ‘ Zwischenkorper’ of the first” (p. 334). All the evidence that my 
study of this stage yields tends to corroborate the correctness of Mathews’s 
observation. Figures 70 and 71 seem to show transition stages of such 
process. The evidence from these figures, coupled with the fact that the 
actual formation and growth of the second spindle at the pole of the first 
has never been seen, renders it very plausible that the second spindle is 
formed as described by Mathews. Furthermore, the central portion of the 
first spindle is never seen to disappear, and after the second spindle is fully 
formed two centrosomes are still often seen in the centrosphere. It is strik- 
ingly characteristic of the second spindle that it usually bends to one side of 
the first polar body, the second being then extruded to one side of the first 
(fig. 74). Such oblique position of the second polar spindle seems to add 
evidence in favor of a rotation from a tangential position, but facts like those 
just recorded render the normally tangential origin of the second polar 
spindle very improbable. The second polar spindle is somewhat shorter and 
slighter in bulk than the first (figs. 71, 72, 73). It again shows very con- 
spicuously a mid-body composed of granules on the spindle fibers in the 
equatorial plane (fig. 73). The second polar body is constricted off along 
the line of the mid-body (fig. 74). It is slightly smaller than the first polar 
body and contains 18 single small dumb-bell-shaped chromosomes. In other 
respects it is similar to the first polar body. The chromosomes soon lose 
their dumb-bell shape and become short, stout rods or spherical masses. At 
this stage the chromosomes of the first polar body have assumed similar 
shapes or occasionally have become massed into two larger clumps con- 
nected by a strand of chromatin (figs. 73, 74). The eighteen dumb-bell- 
shaped chromosomes remaining in the egg after the second polar body is 
formed undergo similar transformations as those in the polar bodies, be- 
coming short, stout, cubical bodies or small spheres (fig. 75). The centro- 
somes have disappeared from view and distinct trace of the centrosphere is 
lost. The astral rays still persist and are often seen to accompany the female 
pronucleus up to the time of fertilization, so that the egg-nucleus seems to be 
provided with an aster of its own, as also reported by Tennent (1906) in 
