700 



BRITTLE RACES OF OENOTHERA LAMARCKIANA. 



the introduction of the barren grains and the mutants into the cal- 

 culation, these figures become changed into those given in table I. 

 It is evident, however, that they indicate equal groups of the two 

 main types. The same conclusion follows from the average of the 

 harvest of the different generations. 



Both scindens and tarda constituted uniform types, which remained 

 the same during all the generations. No external characteristics 

 were discovered on which a distinction of scindens from Lamarckiana 

 could be founded. If we try to express the result of our cultures in a 

 simple formula, giving the constitution of 0. scindens, and call the 

 mutated velutina gametes tarda, we get the formula: 0. scindens = 

 laeta -\- tarda. In this formula our race is assumed to be isogamic 

 like its parent. After self-fertilization it must then split into 50 

 per cent laeta x tarda = scindens, 25 per cent tarda x tarda, and 

 25 per cent laeta x laeta, which last constitute the barren grains. 

 The deviations of our figures from this calculation may be explained 

 by the fact that laeta germs are produced ordinarily in too small 

 numbers, whereas the proportion of velutina is often far above the 

 expectation (8). 



In order to control the exactness of these conclusions, I have 

 made a series of crosses, partly with other species, and partly with 

 0. Lamarckiana and some of its mutants. The first group produced 

 hybrids of the types of laeta and velutina, respectively densa and 

 laxa, which were exactly the same as the corresponding hybrids of 

 the parent species, cultivated beside them for comparison. The results 

 are combined into tables II and III. In the case of 0. scindens x 

 0. Hookeri, the seedlings of crosses on four different individuals 

 were separately counted. There were no essential differences, and only 

 the average figures are given. In the cases of 0. scindens x 0. biennis 

 Chicago, some of the laeta and some of the velutina were self-fertilized 



Table II. 



Crosses of 0. scindens with other species. 



