6 ARKIV FÖR BOTANIK. BAND 17. N:0 12. 



Colombian C. cundinamarcensis bas never been described, 

 and I have not seen any specimens. Hence it seems clear 

 that we must call the Ecuadorian species C. candamarcensis 

 Hook., though I am not convinced that the name is suitable 

 in point of language (and geography!). 



To sum up: Solms in his monographs (1889, 1893) speaks 

 of a monoecious C. cundinamarcensis Hook., native of Ecua- 

 dor, which belongs to sect. Vasconcellea. But the truth is, 

 that the Ecuadorian species should be called C. canda-^ 

 marcensis Hook, which is dioecious and belongs to sect. He- 

 mipapaya. What the Colombian C. cundinamarcensis re- 

 ally is, can only be determined by a study of it in its na- 

 tive country, as the European greenhouse-specimens with 

 that name are nowadays certainly of doubtful origin. 



Returning now to the two parthenocarpic species, de- 

 scribed above, it may be asked whether thej' can be found 

 growing wild anywhere or whether they ^re products of old 

 indian gardening. If the latter should be the case, the hy- 

 pothesis may well be adopted that both have arisen by 

 means of crossing between some other, wild species. However, 

 it must be stated at once that from taxonomical studies 

 no positive evidence can be adduced in favour of such a 

 hypothesis. C. pentagojia is a real Vasconcellea with 5-celled 

 fruit and entire stigmata; C. chrysopetala with 5-celled fruit 

 and bifid stigmata belongs to sect. Hemipapaya as does C. 

 candamarcensis. All three have tubercular seeds with a sar- 

 cotesta that does not fill the cavities between the pro- 

 tuberances, which is the case with sect, Eupapaya. Besides 

 the seeds are much bigger than those of C. Papaya. Thus 

 it is clear that C. Papaya cannot be regarded as an even- 

 tual parent of the two parthenocarpic species, but a certain 

 resemblance between those last-mentioned and C. candamar- 

 censis should not be overlooked, especially as regards the 

 shape and peculiar flavour of the fruits. The difference in 

 the shape of the stigmata is perhaps no valid reason to 

 place C. pentagona in another section of the genus than that 

 of the two other species. My opinion is, that all three are 

 nearly related to one another, but that there seems to 

 be no real ground for regarding the parthenocarpic spe- 

 cies as hj^brid offspring of C. candam,arcensis and some 



