XIV] Maternal Characters in Seeds 263 



One further phenomenon of importance has been dis- 

 covered respecting indent peas. It was said above that 

 the true indents are in coloured coats, but it is obvious that 

 a coloured coat, as such, is not sufficient to mak(* peas 

 indent, for there are many round varieties with coloured 

 coats. It is clear from the experiments of Tschermak (273) 

 and Lock (173, 175), as they point out, and from results of 

 my own, that the properties of the indenting- coats depend on 

 two factors, viz. (i) the indenting factor proper, (2) a requisite 

 pigmentation. For crossing coloured, round varieties with 

 "Victoria," a white variety commonly round, F^ bore F, 

 seeds indent. Lock had a similar result from a round, 

 coloured, native Ceylon pea crossed with Satisfaction 

 (white, wrinkled), but as wrinkledness was introduced, F., 

 seeds were indent and wrinkled. I also by crossing Maple 

 (round, purple flowers) wdth Victoria Marrow, found all /% 

 seeds indent. 



In these cases it is evident that the indentinor factor 

 was introduced in the white-flowered plant, and this, meeting 

 its complementary pigmentation-factor from the coloured 

 side, was able to indent the /% seeds. It is worth noting 

 that ''Victoria," which I take to be our "Victoria Marrow," 

 though generally a round type, produces sporadically a good 

 many seeds with some degree of indentation or pitting*. 

 These (and William I) come nearer to the true indent than 

 those of any white-flowered variety known to me. Probably 

 this is due to an imperfect and occasional manifestation of 

 the powers of the indenting factor, which cannot produce its 

 full effect as the pigment-element is missing ; but when that 

 element is brought in by the cross the compound factor is 

 complete and the seeds are indented. 



An experiment of Tschermak's (273, p. 30, Case 9) 

 contributes one more important and instructive' fact, which 

 gives a clear indication as to the identity of this pigmen- 

 tation-factor. He crossed two round-seeded varieties both 

 having coloured flowe7's, and the F., seeds were indented. 

 Now one of these was an ordinary purple type, but the 

 other was of the pink (salmon-pink) colour well known tor 



* I found by sowing that tlie most indented and tlie roundest give 

 identical results in the next generation. The distinction is lluctuational, 

 and is unaffected by selection. 



