Beer —Studies in Spore Development. II. 72 1 
there seems to - be Instead a general opening out of the substance of the 
chromosomes so that they assume a spongy or porous appearance (Fig. 72). 
In either case the result Is the same, and the material of the chromosomes 
becomes gradually dispersed through the nucleus until it again reaches the 
state of a more or less even reticulum, or assumes the appearance of a cloudy, 
fl occulent precipitate which we have already seen characterizes the * resting' 
nuclei of this plant No definite aggregates of chromatin (prochromosomes) 
can be seen In such a resting nucleus (Fig. 73). It should be mentioned 
that throughout this division the chromosomes, or the bands of concentrating 
chromatic material which precede them, show no distinction of chromatin 
and linin. During the heterotype division the spireme could be very clearly 
seen to be composed of chromatin particles (chromomeres) embedded in 
a clear linin thread, but in the somatic divisions nothing of this sort is 
apparent. 
From what has been said above it will be seen that in Crepis , as in 
Galtonia (examined by Miss Digby), there is no stereotyped plan which is 
Invariably followed by the somatic chromosomes in their development or in 
their passage back to the resting state. Apart from minor variations, how- 
ever, we may say that it is the general rule in this plant for an extremely 
early longitudinal division of the chromosomes or chromosome rudiments to 
provide for the separation of the daughter chromosomes during the anaphase 
of mitosis, and that in consequence of this precocious division a general 
parallelism of parts is often to be seen. In a large number of cases observed 
the daughter chromosomes themselves became centrally vacuolated and 
divided at the telophase of a division. 
During the prophase of the following division it is very usual to find 
the chromosomes reconstructed in two halves, and this dual nature of the 
chromosomes is maintained until the separation of these halves is effected 
at the anaphase. In such cases as this, it does not appear to be straining 
the facts to consider the median division of the daughter chromosomes at the 
telophase of one division as an extremely early preparation for the separation 
of their longitudinal halves at the conclusion of the following division. 
This is quite in accordance with what has already been observed by both 
Miss Digby ( 4 ) and Dr. Fraser and Mr. Snell ( 9 ) in other plants. But I must 
again repeat that there are other Instances in which no such distinct central 
vacuolization of the chromosomes can be detected during the telophase, and 
in which we must assume that their longitudinal division is deferred until 
a later stage. 
The frequent occurrence of parallel filaments and bands during the 
somatic divisions of these plants forms a contrast to what is seen in the 
heterotype division in which such parallel parts are much more rarely 
met with. 
In the preceding pages I have exclusively devoted my attention to the 
3 b 
