Nuclear Division in Fnnkia. 
387 
with the nucleolus '); the members of a pair are hardly ever twisted 
round each other. They next contract considerably, and becoming 
much shorter and thicker, assume also a more periplieral position * 2 ). 
They form short thick loops and it is such a stage as this wkick 
is interpreted by Farmer 3 ) as arising by the breaking olf of various 
loops from a spireme. In badly differentiated preparations, (Figs. 46. 
47. PI. IX), it is impossible to distinguish the aggregations of chro- 
matin in these thick and densely staining objects, but in more 
favourable examples , (Figs. 45. 48 & Text Fig. A II') , a distinction 
between chromatin and linin can still be made 4 ). In such cases the 
longitudinal fission (only seen with difficulty in Figs. 46 — 47) is 
clearly visible. 
VI. The reduced number of chromosomes in Funkia. 
Strasburger 5 ) gives the mean of a large number of countings of 
the somatic chromosomes as twenty-four; the same number, though 
there is some difficulty in fixing the number of the small chromo- 
somes, has been found in the pollen mother cells by Strasburger 
and Miyake and in the present investigation. But Strasburger 
(1906) drew attention to the fact that here the eighteen small chro- 
mosomes are much smaller than any of those in the somatic nuclei 
divisions and Miyake 6 ) has suggested that they are formed perhaps 
by fragmentation of the »anlage«. He thinks that there are some 
grounds for proposing twelve, the usual reduced number in the 
Liliaceae as the reduced number in Funkia. It has been stated 
above 7 ) that it was found impossible to base this tlieory on the ap- 
pearance of a smaller number of prochromosomes in synapsis; but 
for some time I thought there was evidence that twelve Segments 
were first formed during the Segmentation of the spireme, and that 
while six of these gave rise to the long chromosomes, the other six 
broke up to form smaller ones 8 ). It is certain that the eighteen 
small pairs of chromosomes are broken off two or three together 
from the spireme, but the number which is connected in this way 
is not always constant. As however they must be supposed 
1) Cf. Wäger. 1904. 
2) Cf. Miyake. 1906. Fig. 122 PI. V. 
3 ) Farmer & Moore. 1905. 
4 ) Cf. Gregoire & Wygaerts. 1903. p. 7 ft’. 
5 ) Strasburger. 1900. p. 45 & P. J. 1906. p. 17. 
6 ) Miyake. 1. c. 1906. 
7 ; See p. 7. 
8) P. 13. 
