FAULL: DEVELOPMENT OF ASCUS. 
83 
nuclear bodies, which have apparently increased in number. These 
granules very soon move away from the nucleus. At first they are 
scattered throughout the protoplasmic mass (pi. 7, fig. 4) but event¬ 
ually pass to its periphery (pi. 7, fig. 5). Soon after the first spindle 
stage they almost entirely disappear. 
In the meantime the protoplasm has been undergoing changes. 
Simultaneously with the dispersion of the extranuclear bodies the 
protoplasm in the neighborhood of the nucleus becomes hyaline and 
more finely reticulated. This hyaline protoplasm possesses a radiate 
structure with the nucleus as a center (pi. 7, fig. 4). These changes 
are indications that the first nuclear division is near at hand. The 
transforming process is extended to all of the dense protoplasm, 
except a coarsely and irregularly reticulated sheath at the periphery, 
in which the red granules have come to lie. At this stage a centro- 
some and aster make their appearance, as shown in figure 5 (pi. 7) 
in a tangential plane that is parallel to the long axis of the spindle 
that is to follow. The steps leading up to the formation of the 
spindle were not determined. 
Th§ first spindle (pi. 7, fig. 6-13) is generally not as striking as 
those of the two succeeding generations. It is intranuclear, a con¬ 
dition that lasts until a comparatively late anaphase stage, generally 
eccentrically placed and terminated by centrosomes from which 
radiate long rays into the surrounding protoplasm. The rays are 
very coarse and long and are readily demonstrable with either the 
triple stain or with the iron-haematoxylin. It is worth noting that 
they stain 
with the centrosome but that they are outo-rowths of this body is 
open to question. Again there is no more evidence that they are 
retracted to the centrosome at the time of their disappearance, as 
has been suggested, than that they originated therefrom. Indeed, 
instead of being resorbed by the centrosome they appear to fade out 
in the cytoplasm. Moreover, since the latter changes so constantly, 
as we have seen, it is not difficult to believe that the astral rays are 
cytoplasmic modifications. 
It has further been suggested (Harper, ’ 97 ) that in Erysiphe 
communis the number of astral rays may be the same as the num- 
ber of spindle threads and continuous with them. Their existence 
previous to the appearance of the spindle, their separation from the 
spindle by centrosomes of even earlier existence than their own, and 
differently from the centrosome. They are connected 
