Chromosome Numbers of Pas si flora — STOREY 
numbers appears to be effected rather uni- 
formly despite multivalence, however, for 
daughter nuclei were rarely seen to contain 
more than, or fewer than, 18 chromosomes. 
There are extremely few abnormal micro 
spore quartets, and correspondingly few abor- 
tive pollen grains. 
DISCUSSION 
Darlington and Janaki Ammal (1945: 
114) give x = 9 as the basic chromosome 
number of the genus Passiflora. This num- 
ber was doubtless assumed on the basis of 
6 species with 2n = 18 and one species 
which was reported as "2n = 84?.” Deter- 
mination of n = 6 for P. pulcheila indicates 
that not 9 but a lower number probably is 
basic in the genus. With the exception of 
the aberrant species P. foetida (2n = 20), 
all of the species examined, including P. lutea 
(2n = 84? ) , comprise a polyploid series of 
which the rnonoploid number is 6. It is rec- 
ognized that the total of 29 species and vari- 
ant forms for which chromosome numbers 
are known or presumed is but a meager sam- 
ple from a genus with over 400 known spe- 
cies, so it seems not unreasonable to suppose 
that other numbers, both euploid and aneu- 
ploid, may exist among species unreported 
here. 
It is perhaps significant from the breeders’ 
standpoint that all of the horticultural forms 
studied occur in the 2n = 18 group. These 
forms exhibit a fairly high degree of inter- 
specific compatibility, as evidenced by the 
numerous hybrids recorded (Bailey, 1935: 
2487 ) ; but the hybrids themselves are almost 
invariably sterile. Their compatibility sug- 
gests closeness in relationship and, possibly, 
a common origin. 
With an assumed basic number of x = 6 
for the genus, the species in the 2n= 18 
group must be regarded as triploids. Tri- 
ploidy, however, is generally considered to be 
a hindrance to the origin of fertile species 
because of the high order of meiotic irregu- 
41 
larity which it induces. It seems unlikely, 
therefore, that the majority of species exam- 
ined, all of which are fertile, have originated 
through triploidy. An assumed basic number 
of x — 3 would afford a number of mote 
nearly reasonable explanations of the origin 
of 2n = 18 species. These species could then 
be regarded as hexaploids of more or less 
ancient origin. As has occurred in many 
polyploid species of plants, cytological evid- 
ence of genomic composition has disappeared 
in the course of speciation and stabilization 
except in chromosome numbers. 
How speciation within a group could occur 
is illustrated by the mutation from P. eduiis 
to the form flavicarpa. The nature of the 
mutation has not been studied, but when it 
occurred it must have been drastic. The form 
differs from the species in several foliar and 
stem characters. The fruit is yellow instead 
of purple, and the seeds are brown instead 
of black. In addition, the flowering habit has 
been modified from strictly diurnal to partly 
nocturnal, and a barrier of almost complete 
incompatibility seems to have arisen between 
the mutant form and the species from which 
it is presumed to be derived. As mentioned 
earlier, the hybrid origin which was suspected 
by Pope (1935: 11) is not borne out by 
cytological studies. 
P. foetida (2n = 20) and its varieties 
make up a group divergent from the euploid 
series to which all other species examined 
belong. Cytological studies to date have pro- 
vided no clue to its probable origin. The 
possibilities exist either that it arose as a sec- 
ondary polyploid from a 2n = 18 species or 
that it belongs to a second euploid series, per- 
haps with a rnonoploid number of 5, for 
which additional species have yet to be dis- 
covered. If the plants growing in Hawaii are 
truly representative of the species, Janaki 
Ammal’s (loc. cit.) determination of 2n =. 
18 as the chromosome number must be con- 
sidered to be in error. 
