Hclianthus Aiuiuus Syncotylcus. 467 



tricotylous strains. Whether from this a pure, ahiKjst 

 itnvariable syncotylous race can be raised, which would 

 no longer be an intermediate race, but show to this the 

 same relation as that between the anrca varieties and 

 variegated plants (see p. 21), is a point on which we 

 are at present completely in the dark. 



As already stated, I found my race in the year 1887. 

 At that time I had sowed a large quantity of seed of 

 Hclianthus aiuiuus, obtained in exchange from different 

 botanical gardens, and found 18 syncotyls amongst 500 

 seedlings. I planted out these only. They flowered to- 

 gether, but the seeds of each individual were saved, sown 

 and examined separately. In 17 lots the proportions of 

 syncotylous seedlings were distributed between 1 and 

 15%, and in the case of only one of them it was 19%. 

 The latter plant had had its cotyledons fused up to their 

 upper margin and was selected as the basis of my race. 

 The offspring of the other plants were not grown, and 

 moreover from this selected individual only syncotylous 

 seedlings were chosen for further ctiltivation. 



These flowered together and were left to be pollinated 

 by insects. This is apparently necessary in this species, 

 or at any rate in my race of it, because isolated indi- 

 viduals fertilized with their own pollen, either artificially 

 or by humble bees, set no seed. In the later generations 

 the seed has always been collected separately from each 

 plant, and in the following spring sown with a label 

 bearing the number of the parent, and examined. In 

 this way I obtained hereditary values for each single 

 plant. ^ The seeds are large and few plants produce as 

 many as 300 seedlings, so that the values are not so 



^ This rule holds without exception, and therefore 1 shall not 

 mention it in the description of the experiments which are to follow'. 



