278 Beer . — -On the Development of the Spores of Ricci a glauca. 
differentiation with the iron-alum has been carried to the right point. It 
can then be seen that these nucleoli consist of a faintly stained matrix in 
which are embedded a number of intensely black bodies (Fig. 3). 
Compound nucleoli of a similar character occur in the nuclei of the 
Musci. In the spermatogenous cells of Atrichum undulatum the nucleoli 
can be quite clearly seen to consist of a lightly coloured matrix containing 
a number of chromatic particles \ 
In several sporangia I have found that the linin threads tend to become 
more or less massed towards one side of the nuclear cavity (Fig. 7). This 
may correspond to a synapsis stage, although I am not prepared to say how 
far reagents are responsible for its production in the present case. 
Following this condition we find that a much more deeply staining and 
thicker thread has been developed, which traverses the nuclear cavity in a 
number of coils or loops (Fig. 8). This is unquestionably the spirem-thread, 
and it differs radically from the short thread described by Lewis in Riccia 
crystallina 1 2 . The thread can often be followed continuously for a consider- 
able distance, and I believe that it forms an unbroken filament. 
In well stained preparations the spirem-thread shows very beautifully 
an alternation of deeply coloured bodies (chromomeres) with lighter areas 
(Fig. 8). 
It is very probable that the increased amount of chromatic material 
which the thread contains at the spirem stage has been derived from the 
nucleolus and most likely at the expense of the chromatic granules which 
this body encloses. I have unfortunately been unable to find the stages in 
the division of the sporogenous cells which lie between the establishment of 
the spirem and the arrangement of the chromosomes at the equator of the 
spindle (Fig. 10). In the spermatogenous cells of the antheridium, how- 
ever, in which a spirem is also developed, the actual segmentation of this 
thread into the chromosomes could be followed, and it was clearly seen that 
during this process the nucleolus became more and more inconspicuous. 
By the time the chromosomes are fully established the nucleolus has been 
lost sight of altogether. 
Both Garber and Lewis have recorded four chromosomes in Riccia 
(Ricciocarpus) natans, and Lewis found the same number in Riccia crystallina . 
In Riccia glauca the number of chromosomes is higher than this, and I have 
been able to determine with certainty that the reduced number is either 
seven or eight (Figs. 9 and 11). The distribution of the chromosomes to 
1 I should like to take this opportunity of correcting an error which I made in a previous note 
upon ‘ The chromosomes of Funaria hygrometrica ’ (New. Phyt., vol. ii, 1903, p. 166). I there 
stated that the number of chromosomes which appeared in the first division of the spore-mother-cells 
was four. Since this was written I have examined properly fixed material of several mosses 
( Funaria hygrometrica , Atrichum undulatum , Mnium hornum , Polytrichum juniperum ) and I 
have found that in all cases the number of chromosomes is far higher than I formerly supposed. 
2 Compare Lewis’s Plate VII, Fig. 35, with my drawing of this stage. 
