48 



THE AMERICAN NATURALIST [Vol. XL V 



velopment. Nevertheless Winkler was able to make 

 these seeds germinate and the second generation of the 

 plants was reared. 8. gaertnerianum produces fruit only 

 in small numbers but the seeds are perfectly developed 

 and germinate readily, the same being true of 8. proteus. 



Re VERSIONS 



Winkler observed a number of cases where the graft- 

 hybrid reverted to one or other of the parent forms. 

 Similar cases of reversions have been recorded for 

 Cytisus Adami and Cratcego-mespilus. These rever- 

 sions were studied with special care in his first hybrid 

 S. tubingense. In several instances where the plant was 

 cut off below the first lateral bud numerous adventitious 

 shoots arose from the cut surface, and while some of 

 these were pure S. tubingense, others were pure Solanum 

 nigrum, the parent species which is nearer to S. tubin- 

 gense. In a similar manner S. proteus was observed 

 frequently to revert to the tomato, but in no case was 

 there reversion to the nightshade. 



Sometimes spontaneous reversions occur. Thus in S. 

 tubingense the apex of a plant was noted which had sud- 

 denly assumed the characters of 8. nigrum. Winkler 

 gives an excellent photograph of this plant. In other 

 cases shoots of mixed nature were seen, some having the 

 structure of chima?ras, half nightshade and half the 

 hybrid form. In these mixed shoots the inflorescence 

 had flowers of two sorts belonging respectively to the 

 nightshade and to the hybrid. Similar mixed inflores- 

 cences have also been observed in Cytisus Adami. 



The Second Generation 

 In S. proteus and 8. gaertnerianum perfect seed is de- 

 veloped and germinates readily. 8. tubingense which 

 sets fruit freely never has the seed fully developed but 

 as we have already stated Winkler succeeded in germi- 

 nating these seeds and rearing plants from them. He 



