1903 | MITOSIS IN PELLIA 43 
centrosphere is very transitory, not persisting from one nuclear 
division to the next, and appearing only irregularly during the 
division with which it is concerned. Still, this transitory centro- 
sphere is a functional part of the mitotic figure during the first 
two or three divisions. In Pellia, at the fourth division, the 
centrosphere may or may not appear, and in subsequent divisions 
it was only rarely that we could identify the body at all. 
Among the thallophytes, sharply defined centrosomes have 
been described by competent observers who are thoroughly 
familiar with all phases of the centrosome problem. 
In the fungi, judging from Harper’s (16) work on various 
ascomycetes, a centrosome is present during the period of free 
nuclear division in the ascus, when it functions in the formation 
of the spindle. After the period of free nuclear division, the 
centrosome behaves in a very peculiar manner in forming the 
young wall of the ascospore. 
The centrosome has received more attention in the algae than 
in the fungi. In papers by Farmer and Williams (12), and by 
Strasburger (34), centrosomes are described in the oogonia and 
segmenting eggs of Fucus. 
During the early segmentations of the fertilized egg, Stras- 
burger (34) was able to observe the division of the centrosome 
and to trace its continuity from one cell to another. In the 
development of the oogonium, however, no such continuity 
could be recognized. In the large apical cell of Stypocaulon, 
Swingle (37) found that the centrosome divides, giving rise to 
the two centrosomes from which the spindle is developed. He 
was able to recognize the centrosome even during the resting- 
Stage of the nucleus. In the tetraspore mother-cell of Dictyota, 
Mottier (23) found comparatively large and somewhat elongated 
centrosomes. These bodies divide and, at least during divisions 
in the tetraspore mother-cell and in the early divisions of the 
germinating tetraspore, persist from one cell-generation to 
another. They develop asters and play an important part in the 
formation of the spindle. 
Lauterborn (21) figures conspicuous centrosomes in Surirella 
and other diatoms. Karsten (20) also describes centrosomes in 
