o. HEILBORN, CYTOLOGICAL STUDIES ON CARICA . 13 



T^hich of the two alternatives is the right one, or whether 

 erhaps both kinds of seeds are produced, can only be deter- 

 lined by means of experimental work. 



It is desirable from a horticultural point of view that 

 jch work should soon be carried out in the tropics, because 

 ?eds produced by crossing would not, of course, breed true 

 nd, consequently, they would not be suitable for the pur- 

 ose of introducing these species into other countries. Be- 

 des we do not even know as yet whether the seeds of 

 . chrysopetala and C. pentagona really are capable of germi- 

 ation. 



The original purpose of this investigation was to look 

 )r some indication of hybrid origin of C. chrysopetala and 

 . pentagona, as, according to the theory of Ernst (1918), 

 "ossing might be thought of as the probable cause of the 

 irthenocarpy in these species. As stated above, from taxo- 

 omical studies no conclusions can be drawn for or against 

 lis hybrid-hypothesis. Again, as regards cytology, the de- 

 cient conjugation of the chromosomes in the heterotypic 

 ivision would seem to reveal those vestiges of crossing that 

 ere looked for, as non-conjugation has been found as a 

 laracteristic of many hybrids. Conjugation is also lacking 

 I apogamous species, many of which have proved to be 

 icient hybrids. Hence it is obvious that the non-conjuga- 

 on of the chromosomes in C. pentagona may be regarded 

 ? a rather strong proof of the hybrid origin of that species, 

 here is, however, an objection to be made against this 

 'gumentation. As indicated above, the degree of conjuga- 

 on seems to depend more, perhaps, on the phj'siological 

 foperties of the different parts of the ovule, than on any 

 ^brid descent of this species. Besides, there is no constant 

 ilation between the number of bivalents and univalents in 

 le reduction division (indicating the chromosome numbers 

 • the parents), which is the case in hybrids of the />ro5em- 

 iheme (Rosenberg 1909). This is, of course, very natural, 

 5 probably most species of Carica have the same chromo- 

 >me number. Con.sequently it still seems impossible to de- 

 de with certainty whether C. chrysopetala and C. pentagona 

 'e hybrids or not, and the cause of their parthenocarpy 

 ust still bo regarded as rather obscure. There is, there- 



