653 
Beer, — Studies in Spore Development . Ill . 
diverge from one another. This last form of chromosome when viewed 
under unfavourable optical conditions can easily be mistaken for a tetrad of 
spherical bodies lying close together. This fact no doubt explains the 
‘tetrads’ described and figured by Osterhout (see Osterhout ( 5 ), p. 160 and 
Tafel I, Fig. i). 
Ring-shaped chromosomes (Fig. 26, h) are also frequently met with in 
these nuclei. During diakinesis the chromosomes are mostly arranged 
peripherally just below the nuclear membrane, but a few chromosomes 
usually remain closely grouped round the nucleolus as though the concen- 
tration of the nuclear contents which is seen at this point during the 
second contraction had not completely passed away (Fig. 25). 
The surface of the chromosomes is, at this time, often somewhat 
irregular owing to the presence of minute, thorn-like processes upon it. At 
these points very fine filaments can be seen to be attached, and these join 
each bivalent chromosome to other neighbouring ones (Fig. 25) . 
Besides these delicate filaments some flocculent achromatic material 
occurs in the nuclear cavity between the chromosomes (Fig. 25). Each 
nucleus contains either one large nucleolus alone or one large one and 
several smaller ones. 
When the spindle is developed and the nuclear membrane has 
disappeared one or more nucleolar bodies can usually be seen to have been 
extruded into the cytoplasm. These bodies occupy various positions, 
sometimes lying free in the cytoplasm and sometimes caught up among the 
spindle fibres. They usually exhibit no definite relation to the spindle 
poles, but occasionally one may lie in the neighbourhood of one of the poles 
of the multipolar spindle. 
These nucleolar bodies are still quite distinctly visible during the late 
anaphase and the early stages of the telophase (PI. LIII, P"ig. 33). 
In a few of my preparations of diakinesis in Equisetum some of the 
nucleoli appear to be disintegrating within the nuclear cavity. These cases 
suggest that at least some of the flocculent achromatic material in the 
nucleus may be derived from the nucleoli. 
Osterhout ( 5 ) has already investigated the development of the spindle 
in great detail, and my observations, so far as they go on this point, entirely 
confirm his work. A multipolar polyarch spindle is developed in the cyto- 
plasm immediately surrounding the nucleus ; the nuclear membrane dis- 
appears and spindle fibres develop within the former nuclear cavity. These 
intranuclear fibres become attached to the chromosomes and place these in 
connexion with the extranuclear spindle fibres (PI. LII, Fig. 28). The 
dissolution of the nuclear membrane is gradual, and it can often be observed 
that it has disappeared completely at one spot whilst it is still clearly 
visible at another. 
The multipolar polyarch spindle is later converted into a multipolar 
