26o Maternal Characters in Seeds [ch. 



several experiments have also been carried out both by 

 Lock, and by myself assisted by Miss Killby. In the 

 following account only the essential features are related. 



As the type of an *' indent " pea I take the French 

 Purple Sugar pea"^ {Graue Riesen of German seedsmen). Its 

 flowers are purple. The seeds have deep, irregular inden- 

 tations, but with experience of peas it is fairly easy to 

 distinguish these ** indents " from all wrmkled types. As 

 always on plants with coloured flowers, the seed-coats are 

 coloured, being here a grey brown (with some purple spots). 

 In connection with the colour of the seed-coats we meet the 

 first complication. So far as is known, the thoroughly 

 indent type of seed never appears except m coloured coats, 

 but round or wrinkled seeds may exist in coats of all sorts, 



Fig. 36. Upper row : Seeds of Rivet. Lower row : Seeds of Polish 

 wheat. Middle row : /^2 seeds borne by F^ plant, all similar and 

 of intermediate length, though segregation has occurred. (Biften's 

 specimens.) 



whether coloured or uncoloured. As regards the nature 

 of the reserve-materials in these peas Gregory found that 

 the starch grains of indents are like those of round peas 

 (see Fig. 9), being large and simple. 



For simplicity's sake I take first the results of crossing 



"^ This pea has soft ^^ sans parchemiii^'' pods, but the characters of the 

 pods are not involved in the problem discussed in the text. 



