xiv] 



Maternal Characters in Seeds 



259 



This case at first sight does not seem altogether para- 

 doxical, for it may appear natural that the size of the 

 seeds should — at all events as regards the upper limit — be 

 governed by the size of the maternal envelopes. The next 

 group of cases however, though showing that the peculiarity 

 is in some way caused by the maternal tissues, suggests that 

 the influence may be of a more recondite nature. 



Fig. 



35 



Polish wheat, with long glumes, crossed with a «^hort glumed 

 Rwet wheat. F^ has glumes intermediate in length. (BifTen's 

 specimens.) 



B. The Indent Peas. 



This example is that of the Peas {Pisum) known as 

 '' Indent." The crosses of this type of pea give rise to an 

 intricate series of results. For our knowledge of^ these 

 facts we are indebted in the first place to Tschermak* : and 



* Tschermak's account is sometimes difficult to follow, because his 

 terminology does not always show whether he is referring to indent or to 



wrinkled peas. 



17—2 



