142 
H. H. BARTLETT 
secondary hybrids are again intermediate. Clearly we are not con- 
cerned here with a new dominant, in the Mendelian sense, but rather 
with the simultaneous origin of a whole group of non-Mendelizing 
characters, each of which is correlated with the increase in the number 
of chromosomes. 
Among the flowering plants we do not know of any way in which 
the number of chromosomes may be experimentally modified. Among 
the mosses however, the brilliant work of the Belgian investigators 
Elie and Emile Marchal has shown that tetraploid races may be 
produced at will. Their results are of the greatest importance because 
of the light which they throw upon such spontaneous mutations as 
Oenothera gigas. 
It has long been known that moss protonemata might be regener- 
ated from bits of the seta or young capsule ; in the past both Pringsheim 
and Correns have obtained such aposporous protonemata. The 
investigations of the Marchals were undertaken to determine (i) 
whether such protonemata would or would not give rise to moss 
plants with antheridia and archegonia, (2) whether, in case such 
plants were obtained, the gametes would have the 2x chromosome 
number of the normal sporophyte and give rise to a new sporophyte 
with the tetraploid chromosome number, and (3) whether the tetraploid 
races would be like the typical form of the species, or different. It was 
found that diploid gametophytes were obtained in which the vegetative 
cells were larger than in normal gametophytes and the generative 
cells were twice as large. The diploid gametophytes of dioecious 
mosses were absolutely sterile, and synoecious. They could be main- 
tained in culture only by regeneration from pieces of the axis. In the 
case of the monoecious mosses the results were far more interesting, 
for the diploid gametophytes gave rise to tetraploid sporophytes, 
which produced good spores, and in turn reproduced the diploid 
gametophyte. The fertile races thus experimentally obtained were 
named Amhlystegium serpens hivalens and Amblystegium subtile bi- 
valvens. Cytological studies showed that the reproduction of these 
new races was by normal fusion of the diploid gametes. The cytolog- 
ical relations were worked out for Mnium, Bryum, Amblystegium and 
Phascum. In the case of Amblystegium it was found possible to 
double the chromosome number again by regeneration from the 
tetraploid sporophyte, but the tetraploid gametophytes thus obtained 
were completely sterile. The new bivalent race obtained from 
