486 Farmer .• — On Spore- Formation and 
described in the case of the Lilies also obtains here, although 
it is more difficult to follow out the process on account of 
the minuteness of the structures in question in this plant. 
The daughter-chromosomes retreat as open V’s with their 
apices directed polewards, and they aggregate as the two 
daughter-nuclei into two groups placed very closely to the 
cell- plate which forms between them. The cell-plate forms 
a cell-wall which exhibits very remarkable curves, the general 
effect of which is to enable it to meet the walls at the middle 
of the spore-mother-cell at right angles. In fact it appears, 
in this respect, to obey ordinary physical laws, since a com- 
parison instituted between it and the behaviour of a soap film 
introduced into a glass model of the spore-mother-cell which 
I had constructed, revealed a most remarkable degree of 
correspondence between the two cases. 
I am unable to say whether the two daughter-nuclei in 
Pellia enter into a condition of rest before they finally divide 
once more. This does happen in some liverworts, for example 
in Scapania and in L ophocolea ; but if it also occurs in Pellia 
it must be very quickly got over, since I never saw any 
instances of it. In fact the whole process is very rapidly 
passed through, and intermediate stages are in any case 
difficult to get. But when the chromosomes of the second 
division are in the equatorial plate, and lying on the spindle- 
fibres, they resemble exactly the corresponding structures of 
the first mitosis. In this they are in marked contrast with 
Lilies. In the latter plants the second mitosis of the spore- 
mother-cell is strikingly different from the first one in all its 
more obvious features. But it does not therefore follow, in 
the case of Pellia , that a second heterotype division follows on 
the first one, because in this plant the homotype vegetative 
mitosis in the germinating spores are also hardly distinguish- 
able from the heterotype mitosis when both are examined at 
the somewhat late stage in which I found them. The hetero- 
type is, however, clearly recognizable in the younger condition, 
both on account of the very early longitudinal fission of the 
linin, and also by reason of the peculiar forms assumed by 
