378 BOTANICAL GAZETTE [JUNE 
broken apart, but were cut by the knife. It is probable that 
the spindle begins to form rapidly at about this stage, although 
it might already have passed its incipient stage and not be 
detected, if it lies closely applied to the nuclear membrane. 
In the study of Lilium Philadelphicum the writer was unable 
to discover the origin of the radiations which appear around the 
daughter nuclei, but subsequent study of Ranunculus demon- 
strated conclusively that they originate around the poles. Fig. 
456 in Dr. Coulter’s article on Ranunculus (10) was furnished by 
the writer as a good example of this. It is from the endosperm 
of R. multifidus, which is a very favorable object for the study of 
such radiations. In the root tips of Allium Cepa the same origin 
was traced step by step. A comparison of figs. 30, 32, 36, 58, 
59, and 60 will show conclusively the origin of the remarkable 
radiations to be seen in well-prepared material of Erythronium. 
The radiations have their origin from the poles, and only later do 
the daughter nuclei push outward and give to the radiations an 
apparent nuclear origin. The radiations at first appear to be very 
straight and regular ( fig. 36), while later they become more or less 
distorted before they begin to disappear (fig. 60). In favorable 
sections centrosomes are visible, as appear in figs. 32, 32, 36, 55: 
and 59. In the stage represented in fig. 76, the attraction sphere 
appears to form a rather indefinite area from which the radiations 
arise. As to whether these bodies are built up temporarily or 
are permanent, the present study gives no information. In either 
case it is proper to call them centrospheres. At least they are 
the centers for the spindle threads and polar radiations. 
The fate of the nucleolus was not discovered. It is still pres- 
ent at the time of the looped mother skein (fg. 26). In later 
stages, at the beginning of the daughter skein, spherical bodies 
were seen in the cytoplasm which may be extruded nucleoli 
(fig. 59). No figures were seen in the entire study which could 
be interpreted as a synapsis stage. The writer has maintained 
that what is usually called synapsis is a mere artifact which can 
be produced at will by using proper reagents. At the beginning 
of the formation of the spirem, however, the chromatin thread 



