1905] MOORE—SPOROGENESIS IN PALLAVICINIA 87 
FARMER does not discuss the origin of the achromatic spindle, 
evidently regarding that as a matter of minor importance as compared 
with its later behavior in his account of the simultaneous distribution 
of the chromatin. The study of the origin and development of the 
achromatic structure of Pallavicinia Lyellii is attended with consider- 
able difficulties, owing to the large number of chloroplasts in the cell. 
However, it seems to conform in general to the type described by 
Davis (4) for the corresponding phase of Pellia. He finds that 
kinoplasmic caps form over the lobes of the nucleus and extend down 
over it, finally forming fibrillae which enter the nuclear area. In 
my preliminary note (15) I described a similar process for Pallavicinia. 
I found aggregations of kinoplasm at the angles of the nucleus, and 
out of this material fibers are formed, which extend down over the 
protruding portion of the nucleus. FARMER (8) has recognized in 
one of my figures representing this stage the same structure as his 
quadripolar spindle. 
In P. Lyellii this structure is never so prominent as that described 
by Farmer, but his figures do not distinguish clearly the spindle 
fibers from the nucleus. My preparations show a decided lobing of 
the nucleus, but with very slight indications of differentiated fibrillar 
. protoplasm over the lobes. I find no astral rays and no evidence 
whatever of the existence of centrospheres or centrosomes. DAvIs 
(4), CHAMBERLAIN (2), and GrécorrE and WycaAeErts (9) find 
asters and kinoplasmic caps well developed in other periods of ontog- 
eny, but do not find them so prominent, if at all; in the spore mother 
cell. Farmer indeed does not mention the presence of asters’ in 
Pallavicinia, nor does he figure them. Davis (3) in his investigation 
of Anthoceros was the first to question the presence of centrosomes 
in the spore mother cell of liverworts. My studies lead me to hold 
similar doubts and to believe with him that the spindle fibers in the 
spore mother cells of liverworts develop independently of centro- 
somes, so that multipolar stages in spindle formation may be expected, 
as OsterHout, Morrier, and Jvet established in 1897 in the 
pteridophytes and spermatophytes. 
CHAMBERLAIN (2), who studied the germinating spore of Pellia 
with special reference to the centrosome problem, describes a peculiar 
structure in the form of a vesicle fitting over the end of the nucleus, 
