M. H. Crank 7 



in recessive numbers, the family consisting of sixteen pyriforin and 

 four calabash, a straight 8 : 1 ratio. 



One family was constant to full long and oiiu to pyiifonii long. 



Family No. j\ (Plate VII, fig. 19) is a family from a full-long 

 parent, and, as can be seen, there is a slight variation in the shape of 

 till- fruit of individual plants. But such differences are not infrequent. 



It a|)pears that the full longs have two factors which constitute the 

 shoulder of the fruit, and that in the absence of one factor the fruit is 

 pyriform, and in the absence of both it is the calabash type. 



Family No. 12 (Plate VII, fig. 1<S) throws considerable light on 

 the .genetics of fruit shape ; the fruit of the parent was conical and 

 bilocular (as in fruit No. 2 of Fig. 18). A few of the fruits were however 

 round and had three or four cells which, as is previously stated, is 

 characteristic of the F^ plants. This plant gave four distinct shapes — 

 conical, mund, lung, and ])luril<icular compressed fruits. 



AB Ab aB ab 



compressed 

 conical round long round 



Obstrved 29 : G : 14 ; H 



3.51 

 Expectation based on a 9:3:3:1 ratio 'iSIJ'i : 954 : 9-.54 : 3-18 



38-16 



The last type appeared in recessive numbers 8 in .51, or approxi- 

 mately 1 in 16. These fruits are large and somewhat corrugated, like 

 those of many cultivated forms. The plurilocular, large, compressed 

 fruits, as in Fig. 18, No.s. 28, 38, 39, is evidently a recessive appearing 

 as 1 in 16 ; the long members of the family contain B, one of the two 

 factors concerned ; the round contain the other, A ; and the conicals 

 contain both ; but until further experiments are completed, it is not 

 possible to assign the })roperties of these two factors with perfect 

 confidence. 



On the above hypothesis the longs are in excess, but further evidence 

 is required to show if it is of any significance or merely accidental. 



The compressed round fruits always have many cells, and the rounils 

 also, but in a less degree. The compressed round form average about 

 seven cells, and the round three or four, but occasionally they have only 

 two This is characteristic of the original </ parent (Lister's Prolific). 



It appears that the shape of the fruit and the number of loculi 

 are correlated, and that in the absence of the factor B the fruit is 



' It is difficult to grade the rounds from the conii'iil witli certainty- 



