Chromosome Numbers of Pas si flora — STOREY 
39 
On the basis of somatic chromosome num- 
bers which are reported in Table 1 and of 
numbers previously reported in the literature, 
all of the species, hybrids, and varieties in- 
vestigated to date may be classified into six 
chromosome number groups. These groups 
and the species belonging to them are as 
follows: 
2n = 12 — P. pulchella 
2n = 18 — P. mollissirna; P. manicata; P. 
vitifolia; P. quadrangular is ; P. ligularis; P. 
Seemanni; P. maliformis; P. laurifolia; P. 
maliformis X laurifolia; P. racemosa ; P. coc- 
cinea; P. racemosa X coccinea; P. incarnata; 
P. edulis; P. edulis f. flavicarpa; P. alata; P. 
caerulea ; P. alata X caerulea (Pfordti); X 
P. caerulea; P. subpelt at a; (P. foetida ?). 
2n ‘ 20 — P. foetida, and 3 variant forms; 
P. foetida var. gossypifolia. 
2n = 24 — P. suberosa 
2n — 36 — P. suberosa 
2n = 84? — P. lutea 
The 2n = 12 group consists of but a single 
species, P. pulchella. 
The 2n = 18 group comprises 15 species, 
4 interspecific hybrids, and 1 botanical form, 
with 1 species doubtful. Janaki Ammal (loc. 
cit. ) determined the somatic chromosome 
number of P. foetida to be 18. The writer, 
on the other hand, has examined considerable 
material of P. foetida, its botanical variety 
gossypifolia, and several variations, all of 
which occur as common wayside weeds in 
Fig. 1. Mitotic metaphase in P. foetida root tip 
showing 20 somatic chromosomes. 
Hawaii, and has never failed to find 20 
somatic chromosomes (Fig. 1). Examina- 
tions of dividing sporocytes have consistently 
revealed 10 bivalents normally paired (Fig. 
2 ) . The writer, therefore, is disposed to place 
P. foetida in a separate group consisting only 
of itself and its varieties. The possibility is 
not excluded, however, that the Hawaiian 
representatives of the species may be aber- 
rant forms. P. foetida probably was limited 
i i 
Fig. 2. Meiotic metaphase I in P. foetida show- 
ing 10 bivalent chromosomes. 
to a single introduction from which all of 
the plants presently populating the several 
islands are derived. 
Microsporogenesis was observed to pro- 
ceed along a normal course in all of the spe- 
cies in the 2n = 18 group. A single excep- 
tion was noted in P. subpeltata, in which syn- 
diploidy occurred in one locule of an anther. 
The sporocytes in this locule were seen at 
diakinesis to have 2 large nucleoli and 18 
pairs of associated chromosomes instead of 
the usual 9 pairs. There seemed to be no 
strong tendency to form multivalent config- 
urations, and it is supposed that such sporo- 
cytes would have proceeded to give rise to 
diploid microspores. The condition very 
probably arose through failure of a mitotic 
anaphase in the meristem from which the 
sporogenous tissue was derived. 
Pope (1935: 11) proposed a hybrid ori- 
gin for P. edulis f. flavicarpa, possibly as a 
cross between P. edulis and P. ligularis. The 
supposition is not borne out in studies of 
cytological behavior, for meiosis is normal 
in every respect (Fig. 3 ) , and both the ovules 
and the pollen grains are fully viable. The 
