359 
by convergence of bundles of spindle fibres, but convergence is 
not carried so far as in the bipolar types, and the result is a spindle 
with truncated ends. The bundles of fibres end in knob-like en- 
largements, which resemble centrosomes. Similar enlargements 
were observed by Debski at the ends of the spindle-fibres in Chara 
fragilis. 
The general upshot of the various observations in regard to the 
centrosome is that such a body is absent in most forms of plants. 
Strasburger observes that centrosomes can be demonstrated in 
thallophytes and bryophytes, but that in pteridophytes and phan- 
erogams the most careful search for them was in vain. If Oster- 
hout’s and Mottier’s descriptions of spindle formation are correct, 
the conception of the centrosome as a permanent morphological 
element of the cell must be given up, at least so far as these plants 
are concerned. The very interesting observations of Juel add 
further evidence in this direction, Juel found that mitoses some- 
times occur, in which one chromosome becomes isolated either be- 
fore or after division. It forms a small cell by itself, with perfect 
cell-plates between it and the large daughter-nuclei. It acts like 
a cell in all respects ; forms a nuclear membrane; passes into the 
resting state, and even goes so far as to form a complete spindle 
and to divide by mitosis. Such a case shows that a centrosome is 
not a necessary element in mitosis, and one must agree with Juel 
that “those characters which belong to the cell as such are to be 
found not only in the totality of the chromosomes, but also in each 
single chromosome.” 
In several cases the centrosome was very different in form from 
that found in most animal cells. Harper observed that the cen- 
trosome, if present at all in Pesiza and Erysiphe, must be in the form 
of a thin flattened disc stretching across the samewhat blunt pole. 
The centrosome nature of the knob-like thickenings in Chara and 
Basidiobolus is questioned by Debski and Fairchild. 
In regard to the question of chromatin reduction, the most im- 
portant observations were made by Mottier on certain Liliaceae. 
In accordance with these results Strasburger gives up his former 
idea of a purely quantitative division in Lilium by a double longi- — 
tudinal splitting of the chromosomes, and now sustains the view 
accepted by Haecker, Riickert and others that the second division _ 
