Tricotyls as Half Races and Intcruicdiatc Races. 345 



Otis. But hitherto I have not discovered a single in- 

 stance of the former, and have only obtained one in- 

 stance of the latter.^ 



Before I proceed to a detailed description of my 

 races and cultures it seems desirable to give a general 

 account of the manner in which tricotylous seedlings 

 are found, and how the desired half and intermediate 

 races may be most easily derived from them. 



It is well known that amongst the seedlings of dicot- 

 ylous species occasionally individuals are found with 

 three seed leaves. It is only 

 necessary to look over a 

 seed-bed in the garden in 

 spring in order to find in- 

 stances of these. The more 

 extensive the sowing and 

 the more careful our search 

 the greater \n'\\\ be the num- 

 ber of tricotylous seedlings 

 found. Some species pro- 

 duce them in greater, oth- 

 ers in smaller proportions ; 

 and they can often be found 

 even in the smaller pot-cultures of the greenhouse, but 

 in many cases I have had to sow 10,000 or 20,000 seeds 

 of a species before finding a single individual which 

 showed any variation in this direction. But numerous 

 species seem to produce one or several tricotyls in every 

 hundred or thousand seedlings. 



Fig. 62,. Antiryhimim ma jus. A 

 C, D, seedlings with 2, 3. and 

 4 cotyledons. B, with a deeply 

 cleft cotyledon. 



* My Helianthus aninius syncofylcus has not produced a single 

 tricotylous plant in the ten years during which 1 have often counted 

 hundreds or even thousands of seedlings every year. On the other 

 hand they occasionally occur in Helianthus auituus i-arirgalus and 

 some other varieties of the sunflower. 



