Meiosis and Alternation of Generations. 215 
alternating generations, meiosis occurring at tetraspore formation, 
the second great break in vegetative development, and thus a 
rhythmic life-cycle with two alternate, cytologically differentiated 
but vegetatively identical, phases being established. 
But the case of the Floridese with a carposporic phase is not 
so simple. Here we have not only the tetrasporic plant, but the 
immediately preceding carposporic phase, exhibiting the diploid 
condition, so that the carpospores, while they mark a vegetative 
break do not mark a cytological break in development. For this 
reason Svedelius (1911, p. 317) suggests that they should be called 
carpogonidia, and for this reason also Yamanouchi considers that 
the carposporic phase should be considered as part of the sporophyte 
generation, xvhich is continued as the tetrasporic plant. He considers 
that “ the tetrasporic plant may have arisen by a suppression of the 
reduction phenomena in connection with the carpospore, so that it 
germinates with the sporophytic number of chromosomes, producing 
a plant with this number, which consequently becomes at once a 
part of the sporophytic phase. The period of chromosome reduction 
would be thus postponed from the carpospore to a later period in 
connection with the newly formed plant” (1906, p. 435). The 
resemblance of the tetrasporic to the sexual plant Yamanouchi 
explains by the similar environmental conditions under which they 
develop. 
According to this view we can construct a series, starting with 
forms like some of the lower Algae in which meiosis occurs at 
germination of the zygote (Karsten, Spirogyra: Klebahn, Desmids). 
From such forms the ancestors of the Florideae were developed, 
the carposporic phase being intercalated between the zygote and 
the beginning of the next free-living generation, and this intercalated 
phase was characterised by the possession of diploid nuclei, meiosis 
occurring at carpospore formation. This is supposed to be supported 
by the case of Nemalion, in which plant Wolfe concluded, though 
he did not rigorously demonstrate, that meiosis occurred at carpo¬ 
spore formation. Tetraspore formation is absent in this and in 
some other genera of Florideae, and the (supposed) haploid carpo¬ 
spore might be expected to give rise to a sexual haploid plant. 
Finally we have the Florideae, with tetraspores as well as carpospores, 
in which, according to Yamanouchi, meiosis has been deferred from 
carpospore formation to tetraspore formation so that the tetrasporic 
plant retains the diploid character of the “ sporophyte” generation. 
This ingenious theory of Yamanouchi’s, while explaining the 
state of affairs in the higher Florideae, does not fit in well with the 
case of Dictyota, in which there are no carpospores, and in which 
it is quite gratuitous and a priori improbable to consider the 
tetrasporic generation as antithetic. Further we have no evidence, 
beyond the existence of forms with carpospores but no tetraspores, 
that the tetrasporic generation in the higher Florideae arose in 
evolution subsequently to the carposporic phase, nor is there any 
obvious reason why meiosis should have been deferred in the way 
supposed. Taking Dictyota and the Florideae together we have 
three types of life-history in which something more than the sexual 
plant is involved; sexual plant plus tetrasporic plant, sexual plant 
plus carposporic phase, and sexual plant plus carposporic phase plus 
tetrasporic plant. It seems more legitimate, therefore, to consider 
the tetrasporic plant and the carposporic phase separately, rather 
