24 



BULLETIN 1195, U. S. DEPARTMENT OF AGRICX'LTURE. 



The flower clusters which were used for the polhnation experiments 

 were covered with paper sacks just before the first flowers were 

 ready to open. Since the flowers of each cluster opened in succession 

 it was necessary to inspect them at frequent intervals, usually every 

 other day, and pollinate each as it reached the proper stage of maturity, 

 which was considered to be when the terminal pores of the anthers 

 were open ready for dehiscence. As a rule, where self-pollination 

 was employed each flower was polliniited with its own pollen, but in 

 a few instances, where pollen was scarce, that of another flower of 

 the same cluster .was used. In the tests for cross compatibilities 

 pollen from another plant was used. The results are given in Tables 

 4, 5, and 6. 



Table 4. — Results of hand pollinations in Solannni chacoense, in 1921 and 1922. 



Selfed. 



Crosses Between Different Plants Within the Species. 



Every one of the 22 selfed plants list(Ml in Table 4 completely 

 failed to produce seed balls. Oi the .'54 ct)iiibiiiati()ns in crosses only 

 three were successful, and for these every pistil pollinated produced 

 a seed ball with viable seeds. It is, of course, not known how many 

 of these plants were of the same clon, and for only the throe plants 

 which set fruit was it shown that there is a full j)otoncy of pistils. 

 The pollen of all plants available for study was highly potent (see 

 Table 3). It appears, therefore, that tbere are both self-incompati- 



