124 



A. AKERMAN 



found that not all of the outer glumes had the characteristic speltoid 

 structure. On the four lowest spikelets on each side of the rachis the 

 outer glumes were quite typical and distended on the one side of the 

 spikelets, while on the other side they were speltoid i. e. shorter, more 

 abruptly cut off, with more distinct nerves and firmly pressed to the 

 flowering scales (fig. 6). In the upper part of the spike all the outer 

 glumes were of the speltoid type. 



In the autumn 191.7 all the seeds froui the flowers, situated nearest 

 the outer glumes, were sown. Before the sowing the flowers were num- 

 bered, and the seeds were sown in such a manner, that their original 

 position in the spike could be readily determined. In all, 24 seeds 

 were sown. However, only 16 gave rise to fully developed plants. 

 From 7 seeds situated nearest the speltoid outer glumes of the eight 

 lowest spikelets ^ 6 plants were obtained, 5 of which were speltoid, 

 while only one was of the normal type. Seeds from 4 of these spel- 

 toids were sown in the autumn 1918, and from these 102 plants were 

 obtained in all, including plants, both of the normal type and spelloid- 

 heterozygotes identical with the parent plants in a ratio that very well 

 agreed with the ratio 1 : 1 (see table 1.). 



TABLE 1. 



Besides the five speltoid heterozygotes there also came forth one 

 typical plant in the progeny of the flowers inside the speltoid outer 

 glumes in the lower part of the spike. From 64 seeds from this 

 plant 28 plants were obtained all of which were of the normal type. 



From the seeds, situated nearest the typical outer glumes in the & 

 lowest spikelets 5 plants were obtained in all, all of which were typicaL 



From the flowers at the top of the spike, which part morphologi- 



* The eigth was destroyed by the larvae of gall-midges. 



