172 Farmer and Dig by. — Studies in Apospory and 
that the process resembles that of clarissima , Bolton, rather than that of 
clarissima , Jones. 
One of the most striking features of the prothallia of var. unco- 
glomeratum is the very large size of the cells and the nuclei. They are 
far larger than in any of the other varieties of Athyrium with which we 
are acquainted, and the nuclei share with the preceding varieties the exist- 
ence of well-marked and numerous nucleoli (Fig. 31). The antherozoids, 
similarly, are the largest we have met with amongst these varieties of 
A thyrium . 
The number of the chromosomes remains constant through the life- 
history, no reduction taking place in connexion with the gametophyte. 
The actual numbers are very difficult to decide, but we estimate them at 
about 100. They are certainly more numerous than in the other varieties 
we have examined. 
As in the other varieties of Athyrium here dealt with, there is no 
indication of nuclear migration in the prothallium. 
4. Scolopendrium vidgare var. crispum Drummondae . 
This fern was given us by Mr. Druery. It was first recognized as an 
aposporous variety, and as such described by him in 1893 1 . The pro- 
thallia arise in great profusion, especially from the margins of the frond, 
and sometimes they occur even before the frond is pegged down on the 
soil. They are somewhat irregular in shape, but all belong (so far as our 
experience goes) to the expanded type. Antheridia and archegonia are 
freely produced, and are formed on both surfaces of the prothallia, though 
more abundantly upon the lower surface. We have raised young plants 
to test their purity, and find that the variety breeds true on the whole, 
though there is some latitude in the ‘ crisp ’ character of the fronds. 
The chromosomes are very numerous, but they follow what we have 
found to be the rule in all the aposporous ferns which we have studied. 
That is, there is no cyclical alteration in the number of chromosomes ; 
but we found a larger difference in the numbers given by various countings 
than we have met with in most of the other ferns, 2 and we are inclined to 
think these variations probably are not due entirely to errors of estimation, 
but do really correspond to fluctuations in the numbers actually present in 
the different cells. Thus we found in the embryo a considerable number 
of cases in which the chromosomes were about 95-100, but two instances 
in which we could only distinguish about 80. In the prothallial cells 
the average number was 70, whilst in the archegonial nuclei only 80-83 
chromosomes could be identified. The antheridial nuclei are difficult to 
deal with on account of their small size, but in those cases on which we 
1 Journ. Lin. Soc., xxx. 
2 Cf. however the results on pp. 182, 188 for Lastrea pseudo-mas var. cristala apospora. 
