F ei 
1897 | CURRENT LITERATURE 217 
ing the work among seven of his research students. This method of securing 
large results from some single problem, rather than small results from ‘scat- 
tered problems, commends itself to every laboratory in which a group of 
research students may be working. We are pleased to note that of the seven 
collaborators five are American students. . 
Osterhout investigated the spore mother cells of Egudsetum limosum, with 
especial reference to the question of the existence of centrosomes and their 
participation in the process of spindle formation. He gives perhaps the most 
complete series of stages in the development of the spindle that has yet been 
worked out in a vascular plant. The sequence of events is briefly as follows : 
kinoplasmic fibers form (1) a felted layer around the nucleus; (2) they are 
radially placed ; (3) the fibers gather in bundles; (4) the nuclear membrane 
disappears and the fibers come in connection with the linin network and the 
chromosomes ; (5) the fiber bundles are arranged in two groups to form finally 
a bipolar spindle. Centrosomes are not found and could play no part in the 
process as described. The author also figures characteristic tetrad chromo- 
mes, 
Mottier has made further studies on the pollen mother cells of a number 
of lilies and dicotyledons, chiefly with reference to spindle development and 
chromosome reduction. As to the method of spindle formation, his results 
are in substantial agreement with those of Osterhout. e finds in lilies no 
such governing centers in mitosis as were described by Guignard. He shows 
also that the chromosomes in heterotypic division pass through essentially the 
Same stages in the plants studied as have been described by later authors for 
animal nuclei, and argues strongly for the view that the numerical reduction 
of the chromosomes before the heterotypic division is only a pseudo- reduc- 
tion, and that a qualitative division in Weismann’s sense occurs in the second 
division. We must note that a later joint paper by Strasburger and Mottier? 
revises this conclusion, and returns to the doctrine that every mitosis is accom- 
panied by a longitudinal splitting of the chromosomes. 
The formation of small supernumerary pollen grains in the pollen tetrads 
of Hemerocallis was studied by Jue e finds, as Strasburger has described, 
that these grains in every case owe their origin to the isolation of individual 
chromosomes either before or after splitting in the equatorial plate. Such 
single chromosomes form nuclei which function as normal nuclei in every 
respect. If isolated in the first division the small nucleus forms a spindle and 
divides, just as its normal sister nuclei, It is difficult to see how centrosomes 
could be present for these micro- -spindles. Juel also confirms, for Hemerocal- 
lis, the method of spindle development described by Osterhout and Mottier. 
Debski finds in Chara a type of spindle and cell plate formation which 
3 Uber den zweiten Theilungschritt in Pollenmutterzellen. Ber. d. deutsch. Bot- 
Ges. Heft 6. 1897 
