173 
Apogamy in Ferns. 
could at all rely, the numbers obtained varied between 70 and 82. Whilst 
it seems certain, then, that there is no reduction in the accepted sense of 
the term, there is evidence of a slightly larger number of chromosomes 
in the sporophyte as contrasted with the gametophyte. Too much weight 
must not be placed on this, and there is another circumstance which should 
be mentioned in this connexion that seems to us to indicate a possible 
explanation of the irregularities that we have encountered, not only in this, 
but in the other forms. A study of the common Scolopendrmm vulgare of 
the typical form shows that it contains a much smaller number of chromo- 
somes than does the variety. We find that in the spore-mother-cells, when 
they divide, the reduced (meiotic) number of chromosomes is 32, thus 
confirming Stevens’s 1 and Gregory’s 2 statements for this species. The 
number in the premeiotic cells of the sporophyte of the same plant is 
6 4. It is clear then that these must have broken up into smaller units, 
or have been replaced by such, but it is also obvious that this multiplica- 
tion has not resulted in the production of exact multiples in the variety 
of the chromosomes present in the original parent. Thus it is not 
inherently improbable that individual differences may exist, and that the 
different cells may lack uniformity in the matter of their chromosomes. 
Scolopendrium vulgare is remarkable for the number of diverse forms 
to which it has given rise, and we propose to examine these further in 
order to test the hypothesis that the variation in form may be associated 
with fluctuation in the number of chromosomes. It is not improbable that 
Scolopendrium and its variety only represent a comparatively extreme 
example of chromosome-variation, and that the reason why one fails to 
arrive at the same conviction with regard to the existence of variation 
in the preceding ferns may be merely due to the divergence from the 
mean being so small as to fall within what we regard as the limits of error 
in observation. 
The archegonia, as has been already stated, are produced on a fairly 
extensive cushion in great numbers, and they occur on both surfaces of the 
prothallium. Embryos are formed apogamously, and they originate, as in 
Athyrium Filix-foemina var. clarissima , Bolton, from the oospheres. 
The archegonia develop quite normally, and we have been able readily 
to follow all the stages ; these being diagrammatic in the clearness with 
which they can be distinguished. After the ventral canal has been cut off 
from the oosphere the latter body assumes the well-known bun shape, with 
a slight anterior depression. The nucleus is of the usual basin-like form, 
and in the hollow can readily be seen the corpuscle or ‘ body ’ 3 that is 
a regular feature of the normal oospheres. 
1 Stevens, Ber. d. Deutsch. Bot. Gesellsch., 1898. 2 Ann. Bot., xviii, p. 447. 
3 This ‘ body ’ lies sometimes in the ‘ receptive spot ’ ; sometimes, though more rarely, 
outside it. 
