The Autoploids 325 



ot tctraploids, such as leaf size, succulence, ami vitamin content in- 

 crease can be demonstrated, the promise for future polyiiloidy breed- 

 ing offers some hope. Without some initial advantage or promise, the 

 use of polyploidv nnist be questioned for practical purjxjses. 



Direct autotetraploidy in tomatoes has not brought imjirovements. 

 There seem to be hybridization possibilities.-^ Similarly, within the 

 large group of Sohniinn. an interspecific hybridization is probably the 

 most useful aj^proach.-"'' .S. tuberosum, the commonly cultivated 

 species, is already polyploid: doubling is therefore of no value. S. 

 antipoviczii X ^- chticoense amjjhiploid was fertile with S. tuberosum. 

 By this procedine the disease resistance to phytophora from one 

 species, S. antipoviczii, should be transferable into a polyploid hy- 

 brid. ^i^^ The advantages gained from such work can be maintained 

 because vegetative propagation fixed the features once obtained. 



The quality of tetraploid muskmelons, Cucumis meh> I... was 

 definitely superior to the comparaljle diploid variety. ^^ Enough seed 

 can be produced to propagate the tetraploid adequately. These poly- 

 ploids were made in several laboratories; each reported improve- 

 ments. In one instance, taste tests were conducted in such a way that 

 identity of ploidy was not revealed. Without exception, the choice 

 fell to the tetraploid. Since ten different varieties were made tetra- 

 jiioid, a larger number of them were used in comparison Avith the 

 polyploid and diploid. 



A new potential economic species of Cucurbitn Avas developed by 

 doubling the chromosomes of a hybrid between C. maxima and C. 

 moscJiata. One species, C. moschata, carried insect resistance to the 

 hybrid while fruit characters were contributed by the other parent. 

 These characters were not entirely stable in the hybrid, but showed 

 more stability in the polyploid. Fruits matured earlier in the amphi- 

 ploid than in either parent. In the first generation of the amjjhijjloid 

 there was little or no segregation. Later, up to the fifth generation, 

 there appeared segregation for fruit color, shajje, and size. Evidently 

 some intergenomal pairing occinred, and occasional bivalents could 

 be observed during meiosis of the diploid interspecific hybrid. A 

 variant that resembled another species, C. pepo, appeared. This type 

 was completely sterile to either the 2?? or 4/? lines. Since the same 

 variant has reaj^peared, considerable theoretical interest becomes at- 

 tached to this segregate. Large-scale tests in several locations showed 

 that a new jjotential economic species of Cucurbit a has been made 

 (cf. Chapter 12). 



The gigas characters accompanying induced polyploidy became 

 attached to colchicine as soon as the effectiveness of this method was 

 annoimced. Probably the first plantsmen to give serious attention to 

 colchicine were those interested in developing ornamentals. The rea- 



