8 Types of Infiorescence mid Fruits in Tomato 



plurilocular, but that when A is present as in the round fruit, there are 

 not so many cells as when both A and B are absent as in the com 

 pressed rounds. 



At first it appeared that a constant correlation existed between 

 long fruit and the bilocular character, as all the longs which appeared 

 in the F^ were constant to two cells, and some F^ families remained 

 quite constant to this character. 



'I'hree long fruits however occurred with three cells, but only these 

 three were found to have more than two cells after examining several 

 hundred fruits. 



From the results of Family jf, the fruit of the original % parent 

 (Wonder of Italy) apparently possessed the factor B and the original c/ 

 (Lister's Prolific) A. 



Family j'-^ (Plate VII, fig. 19) is a feniily analogous to {|, the 

 zygotic results being : 



A B And Ah : uB : ah 



14 4 1 



Family y^ (Plate VII, fig. 19) is apparently a homozygous conical, 

 and is the only true breeding conical that has at present been isolated. 



The individual fruit ^^ (Fig. 17) appears to be a round that has lost 

 a factor at the shoulder of the fruit, in a similar way to the pyriform 

 longs, and probably it is related to the rounds as the pyriform longs are 

 to the full longs. It is the only plant with that shaped fruit that has 

 appeared, its parent being a F„ conical, and it is curious that more of 

 this type have not appeared'. 



Althiiugh the above work on the inheritance of fruit shape is 

 preliminary, the facts show that many mendelian factors are involved, 

 and doubtless further investigation will assist in elucidating them. 



Ail till' F;; families illustrated in this paper are homozygous for the 

 compound type of inflorescence, and it is possible that some forms of 

 fruit, such as the compressed rounds in family ||, may prove to be of 

 economic value, and theoretically they should be homozygous for 

 fruit shape. 



' In many families containing conicals, rounds and longs, constricted longs occurred, 

 but no constricted rounds or conicals. Further experiment may nevertheless show that 

 some of the constricted types now classed as longs are really constricted conicals. 



