Meiotic Mitoses of Osmundct . 141 
in the daughter chromosomes is very strongly pronounced directly they have 
separated on the equatorial plate (PI. XI, Fig. 95). The chromosomes group 
themselves at the poles (PI. VIII, Fig. ]), and thus the cycle is completed. 
Humphrey ( 12 ) describes ‘ Centrospharen’ in connexion with the division 
figures of O. regalis , but they have not been seen in this investigation. 
Telophase of the Last Archesporial Division. 
In order to interpret the critical and debated phenomena of the hetero- 
type prophases the investigation must be started from the telophase of the 
preceding archesporial division. It is, however, remarkably difficult to 
decide whether the archesporial nuclei of any one particular sporangium are 
in telophase of actually the last division before the onset of the heterotype 
prophase. The growth of the tapetum is not a reliable guide, consequently no 
correlation can be established between the number of thetapetal layers and 
the development of the archesporial tissue, or of the subsequently formed 
spore mother-cells. Matters are further complicated by the rapidity with 
which the telophase is apparently passed through, because exceedingly few 
cases can be found which, after critical examination and weighing of all 
available evidence, can be duly considered to be in this stage. Such 
sporangia have been exclusively selected for this investigation ; the sur- 
rounding sporangia having their nuclei, as a rule, in ‘rest’ or in early 
heterotype prophase. 
Nuclei in late telophase of the last archesporial division are, as would 
be expected, indistinguishable in character from those of the preceding arche- 
sporial mitoses which have already been described (see Text-fig., Nos. 9-12). 
There is the same splitting of chromosomes into two longitudinal halves 
(threads), which may be more or less homogeneous or beaded (Figs. 25 and 
26). As fragmentation of the chromatin proceeds, the halves (threads) 
separate widely, and the chromatic contents become distributed throughout 
the nucleus, and consequently visible chromosome individuality is lost. 
Fine strands join the chromatic portions together (Fig. 27). The nuclei at 
this stage are small and have a definite limiting membrane which causes 
them to stand out sharply from the surrounding cytoplasm. Fig. 28 shows 
a late stage fixed in strong chromic, which accounts for the chromatin of 
the now dispersed chromosomes being in relatively more aggregated portions, 
and staining more deeply— characteristics of this fixative. 
From this stage, and onwards to that of the definite heterotype pro- 
phase, there is considerable difficulty in unravelling the course of events, as 
apparently some of the nuclei pass directly from the telophase of the last 
archesporial division (PI. VIII, Fig. 29) into the heterotype prophase (PI. IX, 
Fig. 35), whilst others pass through an intermediate resting stage (Fig. 34), 
with the characteristic pre- and post-rest phases (see Text-fig., Nos. 13 and 14). 
Consequently it is sometimes impossible to decide whether a nucleus is in 
