1898] KARYOKINESIS IN THE ROOT TIPS OF ALLIUM CEPA 229 
delicate streams of cytoplasm and are of the same nature as 
those shown in fig. 3. They have nothing to do directly with 
the formation of the spindle. The incept of the spindle is very 
sharply differentiated from the surrounding cytoplasm and the 
Space between it and the nuclear membrane appears very clear 
and transparent, like the achromatin of the nucleus. 
After the chromatin band has become considerably thickened 
it loops up into sixteen definite loops, the heads of which, in 
typical cases, point toward the two poles ( figs. 13, 7g). The 
loops, however, do not always seem to have this position in 
relation to the poles, as is shown by figs. 75 and 76. When one 
looks down from one pole nothing is seen of the nuclear spindle 
( fig. 74). The dome-shaped spindle gradually extends out- 
ward and becomes pointed, until the time of the breaking of the 
chromatin coil into a definite number of chromosomes, accom- 
panied by the disappearance of the nuclear membrane ( figs. 
75-20). In these stages the centrospheres become more prom- 
inent, probably through expansion or growth previous to divi- 
sion. The fate of the nucleoli was not discovered. They have 
generally disappeared by the time the chromatin coil has seg- 
mented. In some cases they appear quite vacuolate (fig. 77), 
in others of the same consistency throughout (figs. 75, 76). It 
will be seen from an examination of the figures that the spindle 
is bipolar from the first. It arises as two closely applied caps 
on opposite sides of the nucleus at the summits of which are 
well defined centrospheres. These centrospheres gradually 
extend outwards, drawing the spindle into a sharp pointed 
bipolar structure. In the case of the onion, therefore, it is an 
impossibility for the spindle to arise by an aggregation of many 
cytoplasmic radiations which first form multipolar structures 
passing out on all sides of the nucleus, as has been described 
by Mottier, Osterhout, and others. The spindle is so sharply 
defined from the very first that it can be traced step by step in 
all its Stages of development, its limits always appearing with 
proper staining very distinct and sharply differentiated from the 
cytoplasm. In some cases, where the cells are very flat in 
