57 2 Humphrey. — On some 
spheres to nuclear division, and their apparent control over 
the formation of the spindle in determining the position of its 
poles, are too familiar to need description here. My own 
experience agrees with that of Farmer and of Strasburger in 
regard to the extreme difficulty of recognizing these organs at 
the poles of the spindle in Pellia , or in the resting cells. On 
the other hand, I have observed centrospheres in cells in 
nearly all conditions of rest and of division in a variety of 
plant-tissues, including most of those enumerated above in 
connexion with the investigation of the nucleoli. My ob- 
servations on Psilotiim have been confirmed and extended to 
every phase of cell-life by Guignard (’94), and Strasburger (’95) 
has added Larix europea to the list. 
A favourable plant for the study of the structures in question 
is Osmunda regalis. Here they are not quite as large, as 
a rule, as in Psilotum , but in their occurrence and behaviour 
they quite agree with those of the latter plant. Sections of 
the sporangia of O. regalis , collected at the very beginning of 
growth in the spring, show the various stages in the formation 
of the spores from the spore-mother-cells, often in great 
abundance. But all the mother-cells of a single sporangium 
are in nearly the same phase of rest or division as may be. 
Figs, i to 6 on the accompanying plate show several stages in 
the first division of these cells. Beside the completely resting 
nucleus of the cell shown in Fig. i, lie the very evident 
centrospheres. Here, where they have abundant room, they 
show the usual spherical form, but they sometimes appear 
smaller and flattened in resting cells, when confined in a narrow 
space, as in Fig. 7. Fig. 10 shows the same thing from the 
root-tip of the Onion. In the stage shown in Fig. 3, the 
chromosomes have become individualized, and a remnant of 
the disappearing nucleolus, n , may still be seen. The centro- 
somes have just divided, apparently somewhat prematurely. 
In the two following stages the centrosomes are very plainly 
seen ; but Fig. 6 fails to show them, probably because they 
are hidden by the daughter-nuclei. In the view of the young 
spore-tetrad (Fig. 7), which results from the second divisions in 
