XIV] Maternal Characters in Seeds 261 



indent with wrinkled (e.g. Satisfaction, Laxton's Alpha), 

 for the consequences are there comparatively regular. In- 

 dent fertilised by wrinkled gives the F^ seed indent and 

 unchanged. Wrinkled fertilised by indent gives the A, 

 seed rotmd. This is evidently in consequence of the fact 

 that the wrinkled variety used was a white-llowered plant, 

 with of course uncoloured seed-coats, in which the indenta- 

 tion cannot develop. Such an F^ seed becomes an F^ plant 

 with coloured flowers and coats, and the /% seeds it bears 

 show normal segregation, being indents and wrinkleds* in 

 the usual proportion 3:1. The wrinkled seeds if sown give 

 normal results, viz. 3 wrinkled with coloured flowers : i 

 wrinkled with white flowers. The indents if sown also oive 

 3 coloured-flowered : i white-flowered, but all seeds of 

 ''round" nature on the coloured plants are of course iJidents 

 (w^hether all indents, or 3 indents : i wrinkled as before), 

 while all the corresponding seeds on the white-flowered 

 plants are true rozinds (being either all rounds, or 3 rounds : i 

 wrinkled). 



Thus far all Is intelligible, but when we proceed to the 

 next combination, indent crossed with round, the real diffi- 

 culties are reached. Indent fertilised by a round gives of 

 course the F^ seed indent; and reciprocally, the round ferti- 

 lised by indent gives the F^ seed round. When these seeds 

 are sown, the F-^ plants grow up (with coloured flowers), 

 but the condition of their F^ seeds differs according to the 

 round variety used as original parent. I take first the 

 result which has been most commonly observed. Tschermak 

 observed it in many crosses made with " Victoria," a white- 

 flowered variety, of which more will be said later. Lock 

 saw the same thing with a round-seeded native Ceylon pea 

 having coloured flowers. I have similar results from Fill- 

 basket, Express and Blue Peter, all white-flowered types. 

 In all these cases the F., seeds, borne by F^ plants, are all 

 alike indent, none being round. 



On inspection of such seeds it seems imposslbk? to sup- 

 pose that segregation has occurred. On sowing it is never- 

 theless found that the resulting /% plants are respectively 



* In such a case the sorting oi' living seeds can only be approximate, 

 for to the eye the distinction is not quite sharp. They can be sorted 

 immediately by microscopical examination of the starch. 



